Your SlideShare is downloading. ×
Software for
Technological Patent
Intelligence
e4
Juan CarlosVergara
Alessandro Comai
Joaquín Tena Millán
2006
EMECOM Edic...
Software for Technological
Patent Intelligence
Evaluation of software
and technological intelligence needs.
Juan CarlosVer...
Software for Technological Patent Intelligence
Evaluation of software and technological intelligence needs.
Vergara, Juan ...
TABLE OF CONTENTS
- vi -
- vii -
Table of Contents
Presentation and Acknowledgements
1. Main Findings
2. Introduction
3. Methodology
3.1 Definition...
Software for Technological Patent Intelligence- viii -
5.3 Program evaluated: PatentLabII v1.41
5.4 Program evaluated: PM ...
PRESENTATION AND
ACKNOWLEDGEMENTS
- x -
- xi -
We would like to thank all those individuals and companies who have taken
part in our research, completing the ques...
- xii -
SECTION ONE
Main Findings
- 14 -
- 15 -
Five companies took part and allowed us access to their patent analysis
software, which we were then able to evalua...
Software for Technological Patent Intelligence- 16 -
show that only two of the five software programs cover 50% of these
f...
SECTION TWO
Introduction
- 18 -
- 19 -
2.1 Patent Analysis Software: a review of the situation
A bibliographical review of patent analysis software shows ...
Software for Technological Patent Intelligence- 20 -
2004; Rodriguez, 2003; Lozano, 2003). Other work has focused on tackl...
- 21 -Introduction
- The lack of studies on the demand for software for patent analysis.
That is, the lack of awareness re...
Software for Technological Patent Intelligence- 22 -
in patent analysis. CTI also uses primary and secondary sources speci...
SECTION THREE
Methodology
- 24 -
- 25 -
In order to answer the research questions posed above, two separate sections of
the study were developed, using and...
Software for Technological Patent Intelligence- 26 -
1.- Searching and Downloading
Ability to search in a set of online pa...
- 27 -Methodology
4.- Graphic Generation
Cite Analysis (cited and citing patents in relation to a known patent)
Rankings -...
Software for Technological Patent Intelligence- 28 -
3.1.1 Searching and Downloading
This section assesses all the charact...
- 29 -Methodology
included in the title), is also examined.
Once the indexes have been created, the data included therein ...
Software for Technological Patent Intelligence- 30 -
is related. These links can be shown in a way which is more or less
v...
- 31 -Methodology
3.1.6 Management of Tool
This section cites a group of functions that assess the ease with which
the sof...
Software for Technological Patent Intelligence- 32 -
Criterion
Order
Calculation Selection Criteria Number of
Softwares
1 ...
- 33 -Methodology
We did not, therefore, think it appropriate to include this software program
in our list.
3.2.2 Invitati...
Software for Technological Patent Intelligence- 34 -
relative importance attached by users to the different functions. Thi...
- 35 -Methodology
In our study of demand, we used two databases containing data on professional
individuals and patent use...
Software for Technological Patent Intelligence- 36 -
of which was a total of 102 replies. This final call set a deadline a...
- 37 -Methodology
www.surveymonkey.com).
6
Personal comunication; Juan Manel Batista (ESADE, Barcelona, Spain). To see an
...
- 38 -
SECTION FOUR
Results of the study: demand, users
- 40 -
- 41 -
This chapter gives the results in detail of the poll carried out among patent
users.
Social and demographic informa...
Software for Technological Patent Intelligence- 42 -
In the rest of the sample (43.3%) several sectors are represented, in...
- 43 -Results of the Study: demand, users
Figure 1 – Professions of those polled.
Figure 2 – Sectors represented in the sa...
Software for Technological Patent Intelligence- 44 -
1. Searching in complementary technical / grey literature online data...
- 45 -Results of the Study: demand, users
Often
2. Searching in local (intranet)
databases
0
10
20
30
40
3. Importing pate...
Software for Technological Patent Intelligence- 46 -
1.Automatic patent duplicate detection and removal
2.Automatic groupi...
- 47 -Results of the Study: demand, users
Figure 5 – Filtering and Value Adding.
Software for Technological Patent Intelligence- 48 -
4.6 Local Analysis and Exploitation
This group of functions is charac...
- 49 -Results of the Study: demand, users
Figure 6 – Local Analysis and Exploitation.
2.Automatic abstracts
0
10
20
30
40
...
Software for Technological Patent Intelligence- 50 -
1. Cite analysis (cited and citing patents in relation to a known pat...
- 51 -Results of the Study: demand, users
Figure 7 – Graphic Generation.
Software for Technological Patent Intelligence- 52 -
reports using templates”, were identified as being used “sometimes”.
...
- 53 -Results of the Study: demand, users
Figure 8 –Dissemination and Workgroup.
Software for Technological Patent Intelligence- 54 -
In order to assess this aspect of the software, we have used an appro...
- 55 -Results of the Study: demand, users
Groups Average Correction5
1. Searching and Downloading 2.65 2.25
2. Filtering a...
Software for Technological Patent Intelligence- 56 -
3
Method suggested by Juan Manel Batista of ESADE Business School (Pe...
SECTION FIVE
Comparison of Software: Supply
- 58 -
- 59 -
5. COMPARISON OF SOFTWARE: SUPPLY
In this section the technical specifications for five of the softwares will be
ev...
- 60 -
- 61 -
5.1 Program evaluated: Matheo Analyzer v3.0
Producer: IMCS
8 rue Crillon
13005 Marseille, France
Telephone: +33 (0)...
Software for Technological Patent Intelligence- 62 -
2.- Filtering andValue Adding
Automatic duplicate detection and remov...
- 63 -Comparison of Software: Supply - Matheo Analyzer v3.0
Space or topographic representation of a
patent collection – t...
Software for Technological Patent Intelligence- 64 -
respective fields and give it the appropriate bibliometric treatment....
- 65 -Comparison of Software: Supply - Matheo Analyzer v3.0
Figure 11 – “Matheo Analyzer v3.0” importation.
2) Filtering a...
Software for Technological Patent Intelligence- 66 -
cleaning an index in two very different ways:
a) As a positive filter...
- 67 -Comparison of Software: Supply - Matheo Analyzer v3.0
Thesaurus: Matheo Analyzer does not contain a thesaurus to est...
Software for Technological Patent Intelligence- 68 -
text field with several classifications or descriptors) an analysis b...
- 69 -Comparison of Software: Supply - Matheo Analyzer v3.0
or selected terms from the index) linked to any field.
In this...
Software for Technological Patent Intelligence- 70 -
terms from a particular field (for example, analysis of the relations...
- 71 -Comparison of Software: Supply - Matheo Analyzer v3.0
Analyzer can export matrices to a csv (comma separated values)...
Software for Technological Patent Intelligence- 72 -
Figure 18 – “Matheo Analyzer v3.0” meta-matrix.
6) Management of Tool...
- 73 -Comparison of Software: Supply - Matheo Patent v7.1
5.2 Program evaluated: Matheo Patent v7.1
Producer: IMCS
8 rue C...
Software for Technological Patent Intelligence- 74 -
2.- Filtering andValue Adding
Automatic duplicate detection and remov...
- 75 -Comparison of Software: Supply - Matheo Patent v7.1
Ability to use local databases to integrate
new data and complet...
Software for Technological Patent Intelligence- 76 -
and consultation of patents found, as well as updating them. Finally,...
- 77 -Comparison of Software: Supply - Matheo Patent v7.1
not be repeated. It also permits the selective loading of a spec...
Software for Technological Patent Intelligence- 78 -
9 indices (inventor, applicant, year of priority, year of publication...
- 79 -Comparison of Software: Supply - Matheo Patent v7.1
in MS word, with a variety of options to choose from:
- Quick Re...
Software for Technological Patent Intelligence- 80 -
complete CPIs, applicants/year of publication.
- Short Report: Includ...
- 81 -Comparison of Software: Supply - Matheo Patent v7.1
Figure 23 – The “Matheo Patent v7.1” matrix.
Network: This is a ...
Software for Technological Patent Intelligence- 82 -
Figure 24 – “Matheo Patent v7.1” network.
5) Dissemination and Workgr...
- 83 -Comparison of Software: Supply - PatentLabII v1.41
5.3 Program evaluated: PatentLabII v1.41
Producer: Wisdomain Inc....
Software for Technological Patent Intelligence- 84 -
2.- Filtering andValue Adding
Automatic duplicate detection and remov...
- 85 -Comparison of Software: Supply - PatentLabII v1.41
Ability to use local databases to integrate
new data and complete...
Software for Technological Patent Intelligence- 86 -
and Inpadoc.
5.3.2. Comments on the features studied
1) Searching and...
- 87 -Comparison of Software: Supply - PatentLabII v1.41
a single registry. These operations should be undertaken before d...
Software for Technological Patent Intelligence- 88 -
Figure 27 – Editing a registry-1 in “PatentLabIIv1.41”.
Figure 28 – E...
- 89 -Comparison of Software: Supply - PatentLabII v1.41
3) Local Analysis and Exploitation
PatentLabII v.1.41 has an assi...
Software for Technological Patent Intelligence- 90 -
Figure 29 – 2-dimensional bar graph, “PatentLabIIv.1.41”.
Figure 30 –...
- 91 -Comparison of Software: Supply - PatentLabII v1.41
Thanks to the existence of personalized fields, PatentLabII allow...
Software for Technological Patent Intelligence- 92 -
5) Dissemination and Workgroup
PatentLabII is designed for personal u...
- 93 -Comparison of Software: Supply - PM Manager v1.4.0.3
5.4 Program Evaluated: PM Manager v1.4.0.3
Producer: WIPS Co. L...
Software for Technological Patent Intelligence- 94 -
2.- Filtering andValue Adding
Automatic duplicate detection and remov...
- 95 -Comparison of Software: Supply - PM Manager v1.4.0.3
Ability to use local databases to integrate
new data and comple...
Software for Technological Patent Intelligence- 96 -
PM Manager is able to load files obtained from WIPS Global and then,
...
- 97 -Comparison of Software: Supply - PM Manager v1.4.0.3
Importation: PM Manager can also import registers in electronic...
Software for Technological Patent Intelligence- 98 -
Thesaurus: PM Manager does not permit the creation of lists of key wo...
- 99 -Comparison of Software: Supply - PM Manager v1.4.0.3
Last but not least, with each patent a range of links with docu...
Software for Technological Patent Intelligence- 100 -
Figure 37 – Technology Development Map, “PM Manager v1.4.0.3”.
Figur...
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Vergara comai tena (06) software for technological patent intelligence low
Upcoming SlideShare
Loading in...5
×

Vergara comai tena (06) software for technological patent intelligence low

994

Published on

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
994
On Slideshare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Transcript of "Vergara comai tena (06) software for technological patent intelligence low"

  1. 1. Software for Technological Patent Intelligence e4 Juan CarlosVergara Alessandro Comai Joaquín Tena Millán 2006 EMECOM Ediciones in collaboration with PUZZLE - Revista Hispana de la Inteligencia Competitiva Evaluation of softwares and technological intelligence needs ISBN-10 84-935178-0-1 EMECOM Ediciones in collaboration with PUZZLE - Revista Hispana de la Inteligencia Competitiva Vergara,ComaiyTenaSoftwareforTechnologicalPatentIntelligence EMECOM e4 “What are the main functions used by professionals working in patent analysis? How much importance do users attach to these functions? What software is available for patent analysis and which software meets user requirements most satisfactorily? This pioneering report surveys the needs of intellectual property (IP) professionals, who exploit patents to produce intelligence, and surveys software product capabilitites. Patents are a valuable source of information which, if analyzed, can help to generate knowledge about the relative positions of the different players or establish the state of the art in a given field. The report describes how patent professionals exploit and utilize software packages and it compares to the features of the evaluated software packages. It also shows the value attached to the characteristics provided by the producers. We think that this report, unique in his work, offers a framework for those working with intellectual property” A unique survey of patent analysis software, detailing process steps and examining software features in the context of user need... also potentially useful for future software developers. Martha Matteo, Ph.D. Dr. Matteo is the former director of competitive technical intelligence at Boehringer Ingelheim Pharmaceuticals, Inc (ret.) and currently serves as the vice president for the Society of Competitive Intelligence Professionals (SCIP). Eng - Cubierta Estudio.indd 11/07/2006, 12:281
  2. 2. Software for Technological Patent Intelligence Evaluation of software and technological intelligence needs. Juan CarlosVergara Alessandro Comai Joaquín Tena Millán
  3. 3. Software for Technological Patent Intelligence Evaluation of software and technological intelligence needs. Vergara, Juan Carlos; Comai, Alessandro and Tena Millán, Joaquín Published by: EMECOM Ediciones in collaboration with PUZZLE - Revista Hispana de la Inteligencia Competitiva (www.revista-puzzle.com). EMECOM Consultores, S.L. Llacuna, 162 08018 Barcelona - Spain Teléfono +34 93 401 98 01 info@emecom-ediciones.com http://www.emecom-ediciones.com National book catalogue number: B-35363-2006 ISBN-10 84-935178-0-1 Printed in Spain © Copyright 2006: Juan CarlosVergara, Alessandro Comai and Joaquín Tena Millán. No part of this publication, including cover design, may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronically or optically, without the prior written permission of the publisher.
  4. 4. TABLE OF CONTENTS
  5. 5. - vi -
  6. 6. - vii - Table of Contents Presentation and Acknowledgements 1. Main Findings 2. Introduction 3. Methodology 3.1 Definition of Application Characteristics 3.2 Analysis of supply 3.3 Definition of demand: use, relative needs and value attached to the applications 4. Results of the study: demand, user 4.1 Profession 4.2 Sectors represented 4.3 Experience 4.4 Searching and Downloading 4.5 Filtering and Value adding 4.6 Local Analysis and Exploitation 4.7 Graphic Generation 4.8 Dissemination and Workgroup 4.9 Management of Tool 4.10 Importance 5. Comparison of Software: Supply 5.1 Program evaluated: Matheo Analyzer v3.0 5.2 Program evaluated: Matheo Patent v7.1 vii xi 15 19 25 25 31 33 41 41 41 42 42 44 48 48 50 52 52 59 61 73 TABLE OF CONTENTS
  7. 7. Software for Technological Patent Intelligence- viii - 5.3 Program evaluated: PatentLabII v1.41 5.4 Program evaluated: PM Manager v1.4.0.3 5.5 Program evaluated: Vantage Point v4.0 5.6 Programs not evaluated 6. Conclusion and discussion 6.1 Results of comparison 6.2 Final thoughts 7. References and Autors 8. Annexes 8.1 Annex 1: Letter of invitation sent to software- producing companies 8.2 Annex 2: Letter of Invitation sent to professional individuals 8.3 Annex 3: 3rd. Letter of Invitation sent to professional individuals 8.4 Annex 4: Letter of Invitation sent to professional individuals (in Spanish) 8.5 Annex 5: Functions Table 8.6 Annex 6: Questionnaire 8.7 Annex 7: List of IP Organizations 83 93 105 115 139 139 145 151 157 157 159 160 161 163 165 175
  8. 8. PRESENTATION AND ACKNOWLEDGEMENTS
  9. 9. - x -
  10. 10. - xi - We would like to thank all those individuals and companies who have taken part in our research, completing the questionnaire and attending to our requests for information. We would like to express our gratitude also to all the other individuals and companies we contacted in the course of this study, for the attention and time they have given us and for the interest they have shown. We would also like to thank PUZZLE Magazine for providing us with the space and resources with which to prepare and carry out the poll. PRESENTATION AND ACKNOWLEDGEMENTS
  11. 11. - xii -
  12. 12. SECTION ONE Main Findings
  13. 13. - 14 -
  14. 14. - 15 - Five companies took part and allowed us access to their patent analysis software, which we were then able to evaluate. In order to evaluate the main functions and characteristics of patent analysis software, we prepared a model specifying the concepts into which said functions can be divided. The model was applied to both supply and demand; in other words, our research looked, on the one hand, to experts in patent analysis for their evaluation of the functions and characteristics specified in the model, whilst on the other hand, we also assessed, according to the same model, the software created by individuals or companies to which we had access. We worked on the supposition that the software assessed was representative of the software being offered on the market, although we accepted that our sample was limited and also biased by the eagerness of manufacturers to be included in our study. We concluded that, of all the groups of functions, the section “Searching and Downloading” is the one which adapts least well, in general, to user demand. This weakness is also accentuated by the fact that users gave relative importance a higher rating than patent information searching and downloading compared to the other groups. Only a few functions within this group, such as “Ability to import patent records” for instance, adapt to demand. The group of functions which by and large lives up to user expectations is “Filtering and Value Adding”. In other words, the supply of functions slightly exceeds the use made of them. According to the results of the study, “Local Analysis abd Exploitation” is a group of functions which does not meet user demand. The results obtained 1. MAIN FINDINGS
  15. 15. Software for Technological Patent Intelligence- 16 - show that only two of the five software programs cover 50% of these functions. “Graphic generation” shows positive global results, although there are several areas which are not dealt with quite as persistently as, for instance, graphic and statistical exploitation of the searches carried out. “Space or topographic representation of a patent collection – text mining analysis” or “Ability to use local databases to integrate new data and complete the patent analysis” are hence barely covered by the software studied, when they are in fact used relatively frequently. “Dissemination and Workgroup” is another group of functions which is not given much space in the software analyzed in this study. Alerts, for instance, are not adequately covered despite the fact they are the functions most used by users. There is a major weakness in “Management of Tool”. None of the software included in the study covers the seven functions described for this group in any satisfactory way. This low rating could be due to the fact that the professional individuals who replied are not application administrators (also referred to as webmaster). It is for this reason perhaps, that both the use, as well as the importance, of these functions has a relatively low rating. Overall we have reached the following general conclusions: - No patent-monitoring software fully covers the functions one would expect to find in software of this kind. - Although existing patent-monitoring software meets user demands, there are major gaps in many functions.
  16. 16. SECTION TWO Introduction
  17. 17. - 18 -
  18. 18. - 19 - 2.1 Patent Analysis Software: a review of the situation A bibliographical review of patent analysis software shows that no specialized work has been carried out as regards the evaluation of supply and demand of computer applications for patent analysis. An interesting work was proposed by Trippe 2003, who explored the added value of patent tools available in the market. Up until now, all that existed were a few reports dealing with issues relating mainly to copyright in general (“Patent Tools Survey”)1 . Other minor work has focused on preparing lists of tools available in the market2 . The applications included in this study are used to obtain more advanced knowledge relating to copyright and they are used in Technological Watch or Competitive Technological Intelligence activities. In other words, patents are a valuable source of information which, if analyzed as a group, can help to generate the basic knowledge for creating theories on the relative positions of the different players in a given field. At the same time, the historical study of patents allows future projections to be made, by means of quantitative or statistical tools, or the trends within a sector or a specific company to be identified. Paap (2002) considers, for instance, that the following can be obtained from patent analysis: - The main players - competitors and current and potential collaborators - and their focal areas. - Movement in the interest of the aforementioned players to evaluate the greater or lesser importance given to a technology or a line of research and development. - Organization of the technical endeavors and movement of personnel in time between departments. - Patent strategies used by participants and the opportunities provided and threats posed by the strategies “surrounding the patents”. A number of studies and articles in this field have shown the importance and benefits of carrying out more or less sophisticated patent analysis (Vergara, 2. INTRODUCTION
  19. 19. Software for Technological Patent Intelligence- 20 - 2004; Rodriguez, 2003; Lozano, 2003). Other work has focused on tackling patent exploitation (Paap, 2002; Adams, 2006) or how to organize a systematic collection process using patents as a primary source. As we mentioned earlier, computer applications make the job of statistical analysis or the preparation of patent maps far simpler, thus giving rise to Competitive Technological Intelligence (CTI). It has been shown that computer applications can have a very large number of characteristics and functions aiding the work of experts in this area. For this reason, it is currently highly important that we be aware that applications do exist in the market and that we know which of them can best meet the needs of professional individuals working in CTI. 2.2 Software analysis studies for Competitive Technological Intelligence (CTI) Competitive Technological Intelligence is a practice which specializes in scientific and technological tasks including several types of operations. Generally speaking, the basic process can be compared to the process used by Competitive Intelligence (CI). However, since the emphasis in this context is on technology, CTI uses specialized activities such as Patent Analysis3 (PA). We should emphasize that it is perhaps due to this, that the software designed currently for CI (see for instance: Bouthiller and Shearer, 2003; Nikkel, 2003; Fuld&Company, 2004)4 do not yet include patent analysis, since this is a specialist area belonging to the department of R&D. 2.3 Purpose of the Study This report looks at the supply and demand of software aimed at exploiting patent systems. As we saw earlier in the bibliographical review, no study has been made of how patent users exploit and utilize software packages. In this report, we made our assessment on the basis of two separate studies using the same base structure. Our motives for carrying out this pioneering study can be summarized as follows: - Non-existence of exhaustive studies comparing computer applications for PA.
  20. 20. - 21 -Introduction - The lack of studies on the demand for software for patent analysis. That is, the lack of awareness regarding the use of and the value attached to the characteristics provided by the producers of this type of computer application. - Non-existence of any comparison of applications together with a need expressed by users of said products. - To obtain an assessment of the magnitude and growth of the supply. We recorded over 21 applications existing in the market5 , which in our opinion, is an extensive supply for this specialized field. Another trend which stands out is the increase, if only marginal, in the number of this type of application. - We are dealing with a wide range of available computer applications for PA. An application can have a very large number of functions. However, applications currently existing in the market include different groups of functions and it is, as a result, hard to make any kind of partial comparison of them. For this reason, we believe that it is necessary to standardize or have a uniform approach to the study of these applications in order to make the comparison valid. All of this has led us to formulate several research questions: 1. Which are the main characteristics or functions used by professional individuals working in PA? 2. How much importance do PA users attach to each group of functions? 3. Which software is available for patent PA? 4. Which software meets PA requirements most satisfactorily? The answers to these key questions are given in the following chapters. ********************************** Footnotes 1 See: “PatentCafe’s Patent Software Tools survey” (http://tinyurl.com/96mja) [Consulted on August 11, 2005]. 2 See for instance Paap, J. (2002). Using technical intelligence to drive innovation and technical decisions. Workshop given at the Annual International SCIP conference in Cincinati, USA. 3 It can be observed that the difference between CI and CTI does not occur only
  21. 21. Software for Technological Patent Intelligence- 22 - in patent analysis. CTI also uses primary and secondary sources specializing in recovering technological information. 4 See Assessing Competitive Intelligence Software by Bouthiller and Shearer (2003), Software Report 2004-2005 published by Fuld&Company (2004) (http:// www.fuld.com/Products/ISR2004/HomePage.html) or How can We Determine which Competitive Intelligence Software Is Most Effective? By Nikkel (2003, p.163). Full references can be obtained at the end of the book (see page 151-152). 5 See summary table 1.
  22. 22. SECTION THREE Methodology
  23. 23. - 24 -
  24. 24. - 25 - In order to answer the research questions posed above, two separate sections of the study were developed, using and integrating them in a joint framework. a) Study of the patent CTI software available: this study focused on assessing the different application functions. b) Study of the demand for patent CTI software: this study focused on assessing the subjective needs of users in terms of use and importance attached to the different application functions. Both sections used the same framework for studying the application functions. In other words, the same groups of functions were studied from both a demand as well as a supply viewpoint (see the following section on this). The study of supply was made separately - that is, it was “blind” - with no knowledge of the results of the study on demand. In this way, we tried to avoid any bias in the judgments made in both sections of the study. 3.1 Definition of Application Characteristics In order to evaluate both demand as well as supply, we used a list of software characteristics or functions defined on the basis of: - A review of literature in this field (Ashton y Klavans, 1997; APQC, 2001; Paap, 2002; Trippe, 2003; Dou, et al., 2005; Adamas, 2006). - Analysis of the software available on the market. - and the personal experience of the authors. The functions identified in this way (41 in total) were divided into 6 groups, as shown in the following table1 : 3. METHODOLOGY
  25. 25. Software for Technological Patent Intelligence- 26 - 1.- Searching and Downloading Ability to search in a set of online patent databases Ability to search in other technical/grey literature online databases Ability to search in local (intranet) databases Ability to import patent records Ability to import other records (not patents) Ability to launch simultaneous searches in multiple databases Ability to save search strategies Ability to Schedule repetitive searches Downloading and integration of patent legal status Downloading and integration of graphics Downloading and integration of pdf documents 2.- Filtering and Value Adding Automatic duplicate detection and removal Automatic grouping of patent families Automatic generation of field indexes Ability to define and build new indexes Wizard for grouping and cleaning terms of indexes Patent pertinence (user filled field) Annotation of patents (user filled field) Ability to define and edit patent groups Links to other related documents Taxonomies creation and edition 3.- Local Analysis and Exploitation Automatic extraction of main keywords from patents Automatic abstracts Automatic clustering of patents Automatic classification of patents using semantic filters Full text searching capabilities Semantic searching capabilities
  26. 26. - 27 -Methodology 4.- Graphic Generation Cite Analysis (cited and citing patents in relation to a known patent) Rankings - Analysis of one field. Matrix or Bar graphs – Two field’s co-occurrence analysis Network relations analysis – Two fields co-occurrence analysis Space or topographic representation of a patent collection – text mining analysis Ability to use local databases to integrate new data and complete the patent analysis 5.- Dissemination and Workgroup Ability to publish the contents in the intranet / internet Personalised alerts Alerts to detect changes in the legal status of a patent Automatic reports using templates Ability to export data Ability to create a poll and link a patent to a poll Ability to link a patent to a forum Ability to link a patent to an event in a shared agenda 6.- Management of Tool Ability to publish the contents in the intranet / internet Users access rights management Multi-user access and edition Access and search interface customisation Multilanguage interface Document collections access rights management System utilisation statistics Table 1 - Definition of Application Functions. The following is a more detailed list of the functions.
  27. 27. Software for Technological Patent Intelligence- 28 - 3.1.1 Searching and Downloading This section assesses all the characteristics relating to the process of information collection and its automation. The environment in which a typical user currently carries out his or her work might include access to patent databases as well as access to other bibliographical databases that are normally scientific and technological and which usually complement one another. In addition to this, these databases can be located in a local network (in a private database for instance or in a commercial CDROM-based database) as well as in a website, which means that changes must be implemented in the program in order to allow access to each of these options. Wherever the program included an interface for information searching in a website, we also considered the option for saving the search strategy and for programming its periodic implementation, since these are basic tasks for the Technological Observation function. Another very basic characteristic is the ability to import the results of the searches carried out in any information source, normally in csv format (comma separated values), in text format delimited by fields or in XML format. Lastly, the ability to integrate other information relating to patents in order to add to their value was also assessed. This information is normally in the form of graphics or .pdf documents, but it can also be, for instance: legal information which could increase or remove the value of a patent, or economic information relating to a company or a technology. 3.1.2 Filtering and Value Adding This section covers a whole list of tasks all of which have in common the fact that, when they are carried out, the information becomes far cleaner and better organized and assessed, making subsequent analysis far easier and, in addition, resulting in much firmer conclusions. In patent analysis, it is important to define the information “unit” to be analyzed. Generally speaking, analysts work with “patent families” which group together in one single record all the documents generated from the same priority number. The deletion of duplicate patents and the grouping of patents by families is a task which should be carried out either prior to loading the information into the software or once it has been loaded. Automatic generation of indexes, the ability to easily generate new indexes from elements contained in different information fields (for instance terms
  28. 28. - 29 -Methodology included in the title), is also examined. Once the indexes have been created, the data included therein must be checked for any possible errors and, if necessary, these must be corrected, so that the analyses are correct. The most typical example is the name of an inventor or a company, which may vary where abbreviations are used. Lastly, it is important for users to be able to assess the contents of the patents as they read them. In this way, new information fields are generated which can be analyzed subsequently, such as groups based on specific user interests, links with other valuable documents, or comments on the contents of each patent. Each of these operations adds value to the group of patents to be analyzed, making group work and the making of a final decision on a group of patents easier. 3.1.3 Local Analysis and Exploitation In this section, we assess the basic abilities of the software to manage the information accumulated (filters and advanced searches, classification of results by different criteria, etc.). Other more advanced abilities are also assessed, such as the generation of automatic abstracts for each patent, the extraction of the most representative concepts of each document, the automated “clustering” of the patents into different categories or semantic searching. These abilities already have some relation to text mining. 3.1.4 Graphic Generation This section itemizes the graphics most used by current applications. In general, they can be divided into the following types: - Single-dimension ratings or classifications, normally represented by a line or by bar charts showing one single variable (terms which appear in one field). - Bar graphs or matrixes showing the relationship between two variables (two analyzed fields). These graphics show the number of times two terms appear simultaneously (co-occurrences), each in one information field. - Relationship networks: this graphic allows a fair number of variants. Generally speaking, each term appears as a node in a bi-dimensional space, with lines leading to other terms to which the initial term
  29. 29. Software for Technological Patent Intelligence- 30 - is related. These links can be shown in a way which is more or less vivid or using different colors or associated numbers according to the number of co-occurrences existing between both terms. In addition to this, each node can be larger or smaller or a different color depending on the total number of patents in which it appears. Each node can also be associated to a different icon according to other parameters. Lastly, the position of the nodes in this space and their close proximity to each other may be set by the software in accordance with specific algorithms or they may just float, allowing the user to move them around resulting in a clearer image. - Topographic representations: images in 2 or 3 dimensions which can be complemented by different color tones, showing the more representative concepts, the main classifications or the most important companies in a group of patents. They provide an intuitive vision of the information available in said group and allow analysis to be focused on specific sections of said topographic representation. - Cite analysis: a special kind of representation of relationships between patents, in which the links express the existence of a cite between one patent and another previous one. 3.1.5 Dissemination and Workgroup This section includes a list of the different tasks and functions which can be automated in order to reinforce collaborative work. Firstly, the initial idea is that each user should have his, or her, own information profile and receive any alerts corresponding to said profile. From here, the ability of each user to generate reports using predefined templates in his or her specialty is assessed. In addition, the aim is to generate new knowledge among different individuals by means of discussion and joint analysis of different multidisciplinary issues dealing with said shared information.The possibilities for group work considered were: the generation of polls, the generation of forum discussion and the existence of a shared agenda for the work to be carried out. Several options for exporting information to standard formats, allowing their use in other software, were also listed.
  30. 30. - 31 -Methodology 3.1.6 Management of Tool This section cites a group of functions that assess the ease with which the software adapts to the requirements of different users (for instance: different languages) with different rights (edition or creation, for instance), simultaneous work or whether or not the information can be published in an intranet. The ability to generate system use statistics is included, so that an assessment can be made as to how the use of the software is developing and in short, as to how efficient it is. 3.2 Analysis of supply A review of applications available on the market allowed an initial approach to be put forward with regard to specialized software for patent analysis. 3.2.1 Criteria and selection of patent analysis software The initial software selection was carried out using the following criteria: - the software carries out some kind of analysis. - the supplier is available to deliver a complete copy for assessment. - it must be possible to install the software in the client company’s server.2 - we received a positive answer to our suggestion regarding participation in the study. The following table summarizes the selection criteria, organized in such a way that each criterion eliminates a specific number of applications from our study.
  31. 31. Software for Technological Patent Intelligence- 32 - Criterion Order Calculation Selection Criteria Number of Softwares 1 + Analysis of existing software open to study 33 2 - Those which do not carry out any kind of analysis 12 3 = Those which can be assessed and installed in a server for testing 21 4 - Those which did not receive an invitation from us 2 5 - Those which decided not to join the study 2 6 - Do not confer software license and prefer to bring documentation 1 7 - Those which did not reply to our invitation to take part in the study 11 8 = Assessed software 5 Table 3 – Selection of Patent Analysis Softwares and results of invitation process. By way of an example, we would like to cite two packages which were discarded and which were on the initial list of patent analysis software (see point 1 of the previous table): - BizInt – This package only reformats records obtained from very specific patent databases (Dialog or Questel, for instance). That is, it creates tables and repositions the fields in tables, but it does not carry out any kind of analysis. We did not include this application because the assessment would be pretty well invalid in almost all sections. - Kaliwatch (Pro and Server) is a generic technological observation tool. It has some interesting functions for cooperation between users but it does not focus on analysis, and even less on patent analysis (it does not even mention patents as an information source).
  32. 32. - 33 -Methodology We did not, therefore, think it appropriate to include this software program in our list. 3.2.2 Invitation to participate On the basis of prior analysis, 21 companies were identified as being suitable to partake in the study. They were invited to do so by electronic mail during the first week of July, 2005. The follow-up subsequent to the invitation was carried out by telephone. All of the messages were sent with an acknowledgement of receipt. Very few of these were returned to us and we received a reply stating that the message had been deleted. In addition, the MAPIT mail server produced an error and was not therefore included in our research. Despite the fact that the number of acknowledgements was low, we assumed that a total of 19 companies received our invitation correctly. 3.3 Definition of demand: use, relative needs and value attached to the applications In order to evaluate the situation as regards demand in relation to the characteristics or functions of the software, we carried out a poll among IP professional individuals and users interested in monitoring and analyzing patents and also copyright in general. The questions for which this part of the study hopes to find answers are: 1. Which characteristics or functions do professional individuals use mainly? 2. How much importance do they attach to each group of functions? These two questions establish a starting point for patent analysis software demand. In other words, we wish to identify the potential needs of patent users, without actually specifically defining the real current and future needs of professional individuals. Having carefully pondered the possibilities for studying user needs, we came up with a simple and representative approach. In effect, the approach we suggest in our study focuses on the “use” of the functions described in Table 1 of our poll. In addition, we introduced a section in which we assess the degree of
  33. 33. Software for Technological Patent Intelligence- 34 - relative importance attached by users to the different functions. This exercise allowed us to determine: 1. The degree to which each function is used. 2. The importance attached to each group of functions4 . The main focus of our study is the situation as regards software currently available and whether this software lives up to user expectations. The assessment of user needs and the definition of the future of the software or the determination of the “value” attached to each function incorporated in existing software and included in our sample are not the main focus of this study. Reality does not always match the wishes of users. Despite the fact that a user may wish for a function and value it highly, as long as he has no software program which offers said function, he will have to reply in the questionnaire that he does not use this function. This gives rise to a simplification of the questions. 3.3.1 Instrument The instrument used to collect data was an online questionnaire, accessible on a webpage specializing in information gathering and data processing5 . The questionnaire included 41 functions divided into 6 sections and an additional page where groups were assessed separately (see Annex 8.6). 3.3.2 Assessment of use and relative importance of functions The assessment of the use of the 41 functions (items) divided into 6 groups described in the questionnaire was carried out by requesting information on the following aspects (see Annex 8.5 or page 26): 1. Function use - The degree of use was assessed using a 7-point Likert scale. 2. Relative importance of functions - It was suggested to those responding to the questionnaire that they use an evaluation system in which they awarded 1 point to the factor which was least important to them when compared with the other factors. This method allowed us to define which groups of functions users consider most important6 . 3.3.3 Sample
  34. 34. - 35 -Methodology In our study of demand, we used two databases containing data on professional individuals and patent users. There are currently two main sources amassing the vast majority of professional individuals involved in patent and copyright issues. We discarded other distribution lists such as those originating in the east (see Annex 8.7, for instance) since this project is focused mainly on western populations. These two databases allowed us to reach a significant number of professional individuals. A detailed list of these is given below: - The association PIUG has approximately 600 active members, according to its web page (http://www.piug.org), from 22 countries including the United States of America. The majority of members are from the United States, Europe and Japan. The professional profile of the members includes lawyers specializing in patents, patent agents, people who grant licenses, patent information researchers, patent information salespeople and experts in patent information and documentation. - The mailing list “EPO Mailing List” (http://www.european-patent- office.org/mail.htm). The total number of individuals to be studied is approximately 600 + 800. It should be remembered, that these figures are not exact due to the fact that these distribution lists are voluntary and free and may therefore fluctuate considerably over time7 . In addition, it was impossible to obtain any demographic information on subscribers from either PIUG or EPO since they are anonymous lists open to all8 . 3.3.4 Sending of the invitation The invitation to participate in this initial study was sent by electronic mail. A total of three invitations were sent: 1. On July 11, 2005, the official invitation9 to participate in the poll was sent. This invitation resulted in around 35 replies to the questionnaire. In view of the relatively poor result, we decided to send a second invitation. 2. On July 18, 2005, a second invitation was sent. This produced 53 additional questionnaires. 3. On July 27, 2005, the third and final invitation was sent, the result
  35. 35. Software for Technological Patent Intelligence- 36 - of which was a total of 102 replies. This final call set a deadline at July 31, 2005. Despite the fact that a major increase was observed in the number of replies obtained each time a new invitation was sent, we considered that three invitations gave a sufficiently satisfactory result. 3.3.5 Results The end results of the poll were 102 valid questionnaires. The following table summarizes the results of the PIUG and EPO listing. Ref. Operation Action Number % 1 + Invitations sent 140010 100 2 - Mail deleted 1 - 3 - Unread mail 1 - 4 = Results: net invitations 1398 99.99 5 - Did not reply 1296 92.72 6 = Questionnaires opened 102 7.28 7 - Incomplete questionnaires 0 0 8 = People who replied 102 7.2811 Table 4 – People involved and final sample of users. ******************************************* Footnotes 1 This characteristic is not a selection criteria but it did significantly reduce the number of applications assessed. 2 The profile of those polled and of how the sample was chosen is described in later sections. 3 This part of the poll is relatively simplified. We have taken a simple approach, however, since an assessment of the relevant functions would have become excessively complex otherwise. 4 The scale was as follows: “Not at all”, “Very little”, “Little”, “Sometimes”, “Often”, “Almost every time”, “Always”. “Not Applicable (N/A)” was also added. 5 For further information on the provider used, see Surveymonkey.com (http://
  36. 36. - 37 -Methodology www.surveymonkey.com). 6 Personal comunication; Juan Manel Batista (ESADE, Barcelona, Spain). To see an application which uses this method, see Comai, A (2005) “Factores y Contingencias en la Inteligencia Competitiva: Resultado en un estudio piloto,” PUZZLE - Revista Hispana de la Inteligencia Competitiva, 4(18):12-15. (see http://www.revista- puzzle.com/puzzle_sum_18.htm). 7 By way of an example, it can be observed that the total number of confirmations of having read or rejections of the invitations we sent by electronic mail exceeded 230. It should also be noted that these readings were made long after the questionnaire expired. 8 We contacted both PIUG as well as EPO in order to obtain this information. The replies, however, were negative. In other words, they had no information on the subject. For PIUG, see “http://piug.org/list.html#Majordomo%20Commands%20- %20How%20to%20Join%20the%20Discussion%20List” and for EPO see “http:// www.european-patent-office.org/mail.htm”. 9 See Annex 8.1 10 Estimation (see section Sample, page 35). 11 Approximation (see section Sample, page 35).
  37. 37. - 38 -
  38. 38. SECTION FOUR Results of the study: demand, users
  39. 39. - 40 -
  40. 40. - 41 - This chapter gives the results in detail of the poll carried out among patent users. Social and demographic information This section provides a general overview of the 102 responses we had to the questionnaire. As we mentioned earlier, neither the PIUG nor the EPO list contains any social or demographic information and we have therefore considered it appropriate to include a profile of the IP experts polled. 4.1 Profession There is some variety in the profession of those polled. The questionnaire included three clusters or groups of activity: R+D Manager (6.3%), Librarian (5.3%), and technicians (10.5%).These three activities accounted for less than 23% of the total number of individuals polled but constitute an important minority. The professions identified as most common were “Patent specialist or searcher” (18/102) “Patent attorney” (12/102) and “Copyright manager” (7/102). Nevertheless, lawyers, patent experts and company directors also completed the questionnaire (see Figure 1). 4.2 Sectors represented The sectors represented in this sample are also very varied (see Figure 2). However, with regard to profession, it can be observed that 56% of those polled come from the pharmaceutical, chemical, electronics, computer and engineering and mechanics sectors. There is also a certain homogeneity within this group in which, excepting electronics, the four remaining sectors account for very similar percentages (between 10.3% and 14.4%). 4. RESULTS OF THE STUDY: DEMAND, USERS
  41. 41. Software for Technological Patent Intelligence- 42 - In the rest of the sample (43.3%) several sectors are represented, including consumer goods, biotechnology, cosmetics, software design, as well as consultants, the government and universities. The consultancy group contains 16 companies.1 4.3 Experience We inferred from the sample, that the patent experts have notably extensive experience in everything relating to their work. Despite the fact that we did not specifically ask in the questionnaire whether their experience related exclusively to patents or whether it was wider-ranging, we are inclined to think that their experience in the former is very extensive. In fact, the vast majority of the sample (75.5%) reports having 6 years experience and 53.1%, over 11 years. Figure 3 gives a breakdown of the sample in terms of experience in the field of patents. These results support those obtained for our research, due to the fact that extensive experience in the field of patent analysis is very relevant to an adequate response to our questionnaire. Information on the functions by area of interest This section discusses the results relating to the functions included in the 6 key characteristics of patent analysis software plus a final section analyzing the relative importance of said characteristics. 4.4 Searching and Downloading This characteristic assembles the main patent searching functions in both commercial as well as in private databases. The functions relating to patent downloading/import or search strategy recording in this characteristic were considered at the same time (see Table 5).
  42. 42. - 43 -Results of the Study: demand, users Figure 1 – Professions of those polled. Figure 2 – Sectors represented in the sample. Figure 3 – Years of experience in patent analysis of professional individuals polled.
  43. 43. Software for Technological Patent Intelligence- 44 - 1. Searching in complementary technical / grey literature online databases 2. Searching in local (intranet) databases 3. Importing patent records from other software 4. Launching simultaneous searches in multiple databases 5. Saving search strategies 6. Scheduling repetitive searches 7. Downloading and integration of patent legal status 8. Downloading and integration of graphics 9. Downloading and linking of pdf documents Table 5 – Searching and Downloading functions. Of these functions, the one which is most used (frequent use)2 or which has obtained the highest average rating is “Saving search strategies”. This function stands out from the others due to the fact that around 30% of those polled maintain that they always use it. Other functions obtained similar average ratings, such as “Downloading and linking of pdf documents”, “Launching simultaneous searches in multiple databases” or “Downloading and integration of patent legal status”, for instance, which are used with a frequency very close to “often”. In a second group not far behind the first, we find less frequently used functions. For instance, “Scheduling repetitive searches”, “Searching in complementary technical / grey literature online databases”, “Searching in local (intranet) databases”, “Downloading and integration of patent legal status”, “Importing patent records from other software” are used “Sometimes”. It should be observed, however, that the data for each function is somewhat dispersed.That is, users make quite different use of the functions. A maximum of 32 cases and a minimum of 5 users per type of use were recorded for all functions. Figure 4 shows this distribution. 4.5 Filtering and Value Adding This group of functions, under the heading filters and value adding, allows patents to be managed in a very different way once they have been captured and filed in the company’s database. The functions included in this section are as follows:
  44. 44. - 45 -Results of the Study: demand, users Often 2. Searching in local (intranet) databases 0 10 20 30 40 3. Importing patent records from other software 4. Launching simultaneous searches in multiple databases 0 10 20 30 40 % 5. Saving search strategies 0 10 20 30 40 % % Notatall Verylittle Sometimes 6. Scheduling repetitive searches Little Almosteverytime Always N/A Notatall Verylittle Sometimes Often Little Almosteverytime Always N/A 7. Downloading and integration of patent legal status 0 10 20 30 40 % 8. Downloading and integration of graphics 9. Downloading and linking of pdf documents 0 10 20 30 40 % 1. Searching in complementary technical/grey literature online databases Figure 4 – Searching and Downloading.
  45. 45. Software for Technological Patent Intelligence- 46 - 1.Automatic patent duplicate detection and removal 2.Automatic grouping of patent families 3.Automatic generation of field indexes 4. Definition and building of additional indexes 5. Grouping and cleaning of index terms 6. Evaluation of pertinence (user filled field) 7.Annotation of patents (user filled field) 8. Definition and edition of patent groups 9. Linking to other related documents 10. Creation and edition of taxonomies Table 6 – Filtering and Value Adding functions. The results of the poll show that there is some difference in the way in which these functions are used. Firstly, it should be observed that the most frequently used functions are “Automatic grouping of patent families” and “Automatic patent duplicate detection and removal”. The use associated with these two functions is “often” (4.88 and 4.73 respectively). In addition, 18 and 20 experts in each case stated that they “always” use these functions. On the other hand, the function which is least used is “Creation and edition of taxonomies”. Half of all users of this function report using it “little” (2.46). Other functions, such as “Automatic generation of field indexes” (3.76), “Evaluation of pertinence” (3.70), “Annotation of patents” (3.64) or “Linking to other related documents” (3.60), for instance, have an average use ranging between “sometimes” and “often”, whilst use of the remaining functions, such as “Definition and building of additional indexes” (3.31), “Definition and edition of patent groups” (3.29), or “Grouping and cleaning of index terms” (3.26) appears to be closer to “sometimes”. It should be emphasized, however, that although some functions obtained a specific average rating, responses were also extremely wide ranging. In other words, there are as many experts “always” using a specific function as those “never” using that same function. There are clear differences in the use by experts of function 3) “Automatic generation of field indexes”, shown by the fact that a sizeable number of experts in all use groups appeared on the scale of 7 points we established in the questionnaire. These findings can be seen in figure 5.
  46. 46. - 47 -Results of the Study: demand, users Figure 5 – Filtering and Value Adding.
  47. 47. Software for Technological Patent Intelligence- 48 - 4.6 Local Analysis and Exploitation This group of functions is characterized by its ability to analyze and use patents in accordance with the concepts stated in the following table: 1.Automatic extraction of main keywords from patents 2.Automatic abstracts 3.Automatic clustering of patents 4.Automatic classification of patents in pre-defined categories 5. Full text indexing/searching 6. Semantic indexing/searching 7.Ability to use local databases to integrate new data and complete the patent analysis Table 7 – Local Analysis and Exploitation functions. The results of the study show that the function which stands out most is “Full text indexing/searching” of patents, having ascertained that experts use this function “often”. It should be observed that almost 25% of experts (of the total 82 who answered this question) always use this function. A second group includes the remaining functions with relatively similar ratings, an average use of between “little” and “sometimes). The function in this group which is most used is the obtaining of “Automatic abstracts” (3.73) and the least used is “Automatic classification of patents in pre-defined categories”. The other functions are situated between these two extremes (see Figure 6). 4.7 Graphic Generation The applications used in patent analysis can show the information they are processing in graphic form, allowing an additional, visual exploitation which is somehow far more vivid than in the cases described above. The five functions for which information was requested in this study are as follows:
  48. 48. - 49 -Results of the Study: demand, users Figure 6 – Local Analysis and Exploitation. 2.Automatic abstracts 0 10 20 30 40 3.Automatic clustering of patents 4.Automatic classification of patents in pre-defined categories 0 10 20 30 40 % 5. Full text indexing/searching 0 10 20 30 40 % % Notatall Verylittle Sometimes Often 6. Semantic indexing/searching Little Almosteverytime Always N/A Notatall Verylittle Sometimes Often Little Almosteverytime Always N/A 7.Ability to use local databases to integrate new data and complete the patent analysis 0 10 20 30 40 % 1.Automatic extraction of main keywords from patents
  49. 49. Software for Technological Patent Intelligence- 50 - 1. Cite analysis (cited and citing patents in relation to a known patent) 2. Rankings – Analysis of one field 3. Matrix or bar graphs – Two fields co-occurrence analysis 4. Network relations analysis - Two fields co-occurrence analysis 5. Space or topographic representation of a patent collection – text mining analysis Table 8 – Graphic Generation functions. The two most used functions are “Rankings – Analysis of one field” and “Cite analysis”, with (4.03 and 3.97)2 respectively, suggesting that they are used “sometimes”. These are followed by “Matrix or bar graphs – Two fields co-occurrence analysis” which is used from “little” to “sometimes” (3.59) and “Network relations analysis - Two fields co-occurrence analysis” which is used slightly more than “little” (3.19). Lastly, “Space or topographic representation of a patent collection – text mining analysis” is seldom used, given that the majority of the responses we received fell into the categories “very little” and “little” (2.70) (see Figure 7). 4.8 Dissemination and Workgroup This group contains the following functions: 1. Publish the contents in the intranet / internet. 2. Customized alerts. 3.Alerts with changes on the legal status. 4.Automatic reports using templates. 5. Export all the fields: .csv, .txt, xml, etc 6. Link a patent to a poll with a key question. 7. Link a patent to a forum and begin discussion. 8. Link a patent to an event with a shared agenda. Table 9 – Dissemination and Workgroup function. The function “Customized alerts” (3.99)2 is used “often”. In second place, somewhere between “sometimes” and “often” comes “Export all the fields” and “Publish the contents in the intranet / internet” with 3.69 and 3.57 respectively. Two functions, “Alerts with changes on the legal status” and “Automatic
  50. 50. - 51 -Results of the Study: demand, users Figure 7 – Graphic Generation.
  51. 51. Software for Technological Patent Intelligence- 52 - reports using templates”, were identified as being used “sometimes”. Lastly, another three functions are used “little”. It should be noted that these functions “Link a patent to an event with a shared agenda”, “Link a patent to a poll with a key question” and “Link a patent to a forum and begin discussion” are the ones which were rated lowest in this group, since over 50% of those polled (83) said that they do not use these functions (see Figure 8). 4.9. Management of Tool This group contains the following functions: 1. Management of users access rights 2. Management of Document collections access rights 3. Simultaneous multi-user access and edition 4. Customization of access and search interface 5. Multilanguage interface 6. System utilization statistics Table 10 – Management of Tool functions. The functions above refer to the ability of the software to manage the applications in such a way that user needs are met. It should be emphasized that the overall results obtained in this section are among the lowest in the poll. As is shown in figure 9, user responses are fairly similar for all functions, with an average use which is close to “sometimes”. Although the responses are quite varied, the one which appears most frequently (from 24% to 38%) is “never”. The function which stands out as being least used is “Multilanguage interface”. 4.10. Importance The importance attached to each of the functions included in this section is a relevant estimate of how highly they are rated. This is how we distinguish frequency of use from the relative value each user considers has been added to his work by each group of functions.
  52. 52. - 53 -Results of the Study: demand, users Figure 8 –Dissemination and Workgroup.
  53. 53. Software for Technological Patent Intelligence- 54 - In order to assess this aspect of the software, we have used an approach which relates all the functions to the one given the lowest rating by users3 . In this way, a comparison between all the functions can be established on a qualitative basis which is appropriate for our study4 . A total of 79 experts completed this section. In order to make the evaluation process easier, the questionnaire suggested an assessment scale of 1 to 3 with intervals of 0.25 points. The results of the average user ratings obtained are shown in table 11. The table shows the average of the results obtained and the correction carried out in order to obtain the relative value. Figure 9 – Management of Tool.
  54. 54. - 55 -Results of the Study: demand, users Groups Average Correction5 1. Searching and Downloading 2.65 2.25 2. Filtering andValue Adding 2.27 1.87 3. Local Analysis and Exploitation 1.86 1.46 4. Graphic Generation 1.59 1.19 5. Dissemination and Workgroup 1.59 1.19 6. Management of Tool 1.40 1 Table 11 – Relative importance of functions. Firstly, “Searching and Downloading” is the characteristic most appreciated by users for analyzing patents. This group was rated highest, with results which reflect the fact that those polled attach an importance to it which is more than twice that attributed to “Management of Tool”. In addition to this, more than 65% of those polled gave this type of function maximum rating. The second item, “Filtering and value adding” also stands out as being appreciated almost twice as much by those polled as the reference characteristic. This means that the functions associated with this characteristic rank highly among users. Nevertheless, only 28% of those polled expressed an opinion on this concept and so it definitely has less of an impact than in the case of the previous concept (“Searching”). “Local Analysis and Exploitation” was rated at 1.46. Lastly, another two characteristics were given a similar degree of importance. Those polled rated “Graphic Generation” and “Dissemination and Workgroup” only just above “Management Tool”. The data obtained allows priorities to be established, emanating from the opinions of users regarding which functions require more attention in terms of design and the improvement of the type of software being studied. ******************************************************** Footnotes 1 For instance: legal, copyright management consultants, analysis and assessment of the financial risk in copyright, lawyers and consultants or IP consultants. 2 The scale used was: (1) Not at all, (2) Very little, (3) Little, (4) Sometimes (5) Often, (6) Almost every time, (7) Always y (0) N/A.
  55. 55. Software for Technological Patent Intelligence- 56 - 3 Method suggested by Juan Manel Batista of ESADE Business School (Personal Communication). 4 The questionnaire contains the following explanation for users interviewed: - NOTE: Assign “1” to the least important group of methods/techniques and rate ALL the others groups against it. 6 If you rate “1” it means that the group of methods are equivalent to the one being compared with. If you rate 1.5 then it means that the group of methods are 50% more important to the one being compared with. If you rate 2, that means that it is one time more important or dobble and if you rate 3 it is two time more important and so on... - 5 The correction was carried out as follows: The lowest average rating was identified. In the case of “Software Management” it is 1.40. Since the system takes 1 as the lowest rating, we made the lowest rating the reference rating. That is, reducing 1.4 by 0.4. In this case, “Software Management” became the reference and thus took on the rating 1. The ratings of other items were reduced by the same amount in order to study the incremental ratings in terms of importance, in accordance with the evaluation system used in the questionnaire.
  56. 56. SECTION FIVE Comparison of Software: Supply
  57. 57. - 58 -
  58. 58. - 59 - 5. COMPARISON OF SOFTWARE: SUPPLY In this section the technical specifications for five of the softwares will be evaluated in depth in this study. However, information will also be added on a further ten softwares which have not been fully evaluated. The partial or complete description of the programs is carried out using the format below: - Name of the program and the company’s details. - Evaluation table summarising feature according to the system adopted by this study. - Description and details of the program’s features in six key areas. The technical details of each program are presented in the following section in alphabetical order.
  59. 59. - 60 -
  60. 60. - 61 - 5.1 Program evaluated: Matheo Analyzer v3.0 Producer: IMCS 8 rue Crillon 13005 Marseille, France Telephone: +33 (0)491 082 882 Fax: +33 (0) 491 783 906 E-mail: info@imcsline.com Website: http://www.matheo-software.com Evaluation Table1 MATHEO ANALYZER v3.0 Evaluation 1 2 3 4 5 1.- Searching and Downloading Ability to search in a set of online patent databases Ability to search in other technical/grey literature online databases Ability to search in local (intranet) databases Ability to import patent records Ability to import other records (not patents) Ability to launch simultaneous searches in multiple databases Ability to save search strategies Ability to Schedule repetitive searches Downloading and integration of patent legal status Downloading and integration of graphics Downloading and integration of pdf documents
  61. 61. Software for Technological Patent Intelligence- 62 - 2.- Filtering andValue Adding Automatic duplicate detection and removal Automatic grouping of patent families Automatic generation of field indexes Ability to define and build new indexes Wizard for grouping and cleaning terms of indexes Patent pertinence (user filled field) Annotation of patents (user filled field) Ability to define and edit patent groups Links to other related documents Taxonomies creation and edition 3.- Local Analysis and Exploitation Automatic extraction of main keywords from patents Automatic abstracts Automatic clustering of patents Automatic classification of patents using semantic filters Full text searching capabilities Semantic searching capabilities 4.- Graphic Generation Cite Analysis (cited and citing patents in relation to a known patent) Rankings - Analysis of one field. Matrix or Bar graphs – Two field’s co- occurrence analysis. Network relations analysis – Two fields co- occurrence analysis
  62. 62. - 63 -Comparison of Software: Supply - Matheo Analyzer v3.0 Space or topographic representation of a patent collection – text mining analysis Ability to use local databases to integrate new data and complete the patent analysis 5.- Dissemination and Workgroup Ability to publish the contents in the intranet / internet Personalised alerts Alerts to detect changes in the legal status of a patent Automatic reports using templates Ability to export data Ability to create a poll and link a patent to a poll Ability to link a patent to a forum Ability to link a patent to an event in a shared agenda 6.- Management ofTool Management of users access right Management of Document collections access rights Simultaneous multi-user access and edition Customization of access and search interface Multilanguage interface System utilization statistic Table 12 – Benchmark for the “Matheo Analyzer v3.0”. 5.1.1 Definition of the software Matheo Analyzer is a tool specialising in graphically visualising and analysing information retrieved from bibliographic databases. The purpose is to analyse all types of bibliographic references, retrieve this information from its
  63. 63. Software for Technological Patent Intelligence- 64 - respective fields and give it the appropriate bibliometric treatment. Figure 10 – “Matheo Analyzer v3.0” main screen. 5.1.2 Comments on the features studied 1) Searching and Downloading Connection to databases: Matheo Analyzer does not have modules for carrying out database searches. Its starting points are the lists in text form which contain registers obtained from databases or other applications. For instance, it can use data accessed from exported Matheo Patent bibliographic files. Matheo Analyzer contains an assistant to take the inexperienced user step-by-step through the importation of a list of files. When the process has finished, the correct steps for importing from such a source can be saved for later reference and use. Matheo Analyzer allows the user to carry out various types of importation at any time: for a given project new fields, which had not been imported previously, may be entered. In addition, The program can import new records differentially (records with “key” fields differing from those already loaded can be imported). Matheo Analyzer is not designed to import or manage either graphics or attached pdf documents.
  64. 64. - 65 -Comparison of Software: Supply - Matheo Analyzer v3.0 Figure 11 – “Matheo Analyzer v3.0” importation. 2) Filtering and Value Adding Identification of duplicates: Matheo Analyzer detects duplicated patents by identifying a “key” field which cannot be copied. Working with fields: Matheo Analyzer can create new subfields from existing ones and can create new linked indices. This operation is generally carried out in order to work later with subgroups of special interest; for example, the 10 main patentees of the first 5 classifications. The selection of the terms which a user wishes to incorporate into a new subfield may be defined using various criteria: range of frequency, search using key terms, or direct selection from an index. Standardisation and cleaning of indices: Matheo Analyzer automatically generates as many indices as there are fields defined in the importation process. The user may search through these and examine the patents classified with each term. All fields can be edited. There are two methods to facilitate the creation of a particular field: - Reference table (filtered): This consists of a tool for automatically deleting all the unwanted terms in a particular field. It is made up of a list of terms (or common expressions). This table is of interest for
  65. 65. Software for Technological Patent Intelligence- 66 - cleaning an index in two very different ways: a) As a positive filter, if we are only interested in working with specified terms. b) As a negative filter, if we are interested in deleting specified terms. In this case, the reference table acts as a list of empty words. The reference table can either be entered by hand or by selecting a text file with a list of the terms (one per line). - Correspondence table: this consists of a tool used for automatically “standardising” terms in a particular field. The mechanism used is that of search and automatic substitution based on a double list of terms. The first list indicates the “non-standard” term while the second indicates the “standard” one which should substitute it. This table can be loaded and edited manually but the user can also create a text file, with a pair of terms on each line, and then load it automatically. In both cases, the user can create as many crossed or reference tables (crossed tables) as they wish. Figure 12 – “Matheo Analyzer v3.0” reference table.
  66. 66. - 67 -Comparison of Software: Supply - Matheo Analyzer v3.0 Thesaurus: Matheo Analyzer does not contain a thesaurus to establish equivalence among terms. Text-mining: Matheo Analyzer is not equipped with text-mining technology. Text fields, along with the title, a summary or various search requests, can be loaded, analysed or removed using the reference tables, but they are not analysed using semantic algorhythms. User classification: Matheo Analyzer is not designed so that the user can classify information. It is assumed that this task has been carried out previously. 3) Local Analysis and Exploitation Matheo Analyzer contains three basic elements for exploiting information effectively: - Forms: By selecting a field, the index in that field is displayed with all the terms contained within, and their frequency. This index forms the basis for carrying out other operations. - Pairs: By selecting any two fields, an index with all combinations of pairs of terms and their frequency of co-occurrence is shown. FFigure 13 – A “Matheo Analyzer v3.0” cluster. - Clusters: By selecting a field (which, in most cases, will be a complex
  67. 67. Software for Technological Patent Intelligence- 68 - text field with several classifications or descriptors) an analysis based on the K-means algorhythm is carried out. This classifies the patents into groups sharing certain features, which, in turn, differentiate them from the rest of the groups. 4) Graphic Generation Matheo Analyzer 3.0 can create the following graph types: Histograms: - Frequency histogram: this analyses the content of a field. The height of each bar represents the number of patents corresponding to each term. This is the most commonly used type. - Range histogram: this analyses the frequency of the terms used in a particular field. The height of each bar represents the number of terms in this field with a determined frequency. - Indexing depth histogram: this analyses the lists with a defined number of terms in a particular field. It indicates to what extent this field is wide-ranging (many terms used to define this field) or extremely concentrated (very few terms used to define this field). Figure 14 – “Matheo Analyzer v3.0” frequency histogram. In all three cases a previous condition can be created (based on text, frequency
  68. 68. - 69 -Comparison of Software: Supply - Matheo Analyzer v3.0 or selected terms from the index) linked to any field. In this case, the histogram (from any field) will only analyse the terms within the limits of that previous condition. Network: a two-dimensional cartographic diagram of differing elements (network nodes) and the relationships among them (links among nodes). Each element usually has an associated number indicating its frequency, and may also have links to any other element. In this case, the two elements form a “pair”. This pair also has an associated number indicating the frequency of the relationship between the two elements. By way of a contextual menu, every node in the network allows the visualisation of the patents included within it, as well as allowing the insertion of commentaries in these lists. There are four kinds of network graphics: - Symmetric network: this corresponds with the analysis of the terms in a particular field (for instance, analysis of the co-operative relationship between companies or authors). - Asymmetric network: this corresponds with the analysis of the terms from two fields (for instance, analysis of the relationship between companies and technical fields). Figure 15 – “Matheo Analyzer v3.0” asymmetric network. - Condorcet network: this corresponds with the analysis of the relationships among a group of patents, as they share a group of
  69. 69. Software for Technological Patent Intelligence- 70 - terms from a particular field (for example, analysis of the relationships among various groups of patents). - Propagation network: This corresponds to the maximum deployment of the Relationships among terms in the same field, based on one or more recognised terms (for example, the relationships among inventors, from “Li Ming” onwards). Figure 16 – “Matheo analyzer v3.0” propagation network. Matrix: This carries out an analysis of co-occurrences between two lists of terms. The result is a matrix of cells in which the number of co-occurrences appearing for each one is given. The greater the number of co-occurrences, the darker coloured the cells. They can be of two kinds: - Simple: These can be binary, condorcet, symmetric and asymmetric. They consist of tables whose rows and columns contain the terms in each field or subfield (see figure 17). - Evolved (MetaMatrix): This type of matrix contains the terms of a field or subfield in its columns, but in its rows it contains groups of terms (see figure 18). 5) Dissemination and Workgroup Matheo Analyzer software is for individual use and does not permit groups of users to interact. Neither is it possible to publish information on the web for other users to search for information from another computer. Matheo
  70. 70. - 71 -Comparison of Software: Supply - Matheo Analyzer v3.0 Analyzer can export matrices to a csv (comma separated values) formatted text document readable by any spreadsheet or database software. Figure 17 – “Matheo Analyzer v3.0” asymmetric matrix.
  71. 71. Software for Technological Patent Intelligence- 72 - Figure 18 – “Matheo Analyzer v3.0” meta-matrix. 6) Management of Tool The functions described in this section are applicable to software utilised by various people. These functions are not available in Matheo Analyzer.
  72. 72. - 73 -Comparison of Software: Supply - Matheo Patent v7.1 5.2 Program evaluated: Matheo Patent v7.1 Producer: IMCS 8 rue Crillon 13005 Marseille, France Telephone: +33 (0)491 082 882 Fax: +33 (0) 491 783 906 E-mail: info@imcsline.com Website: http://www.matheo-software.com MATHEO PATENT v.7.1 Evaluation 1 2 3 4 5 1.- Searching and Downloading Ability to search in a set of online patent databases Ability to search in other technical/grey literature online databases Ability to search in local (intranet) databases Ability to import patent records Ability to import other records (not patents) Ability to launch simultaneous searches in multiple databases Ability to save search strategies Ability to Schedule repetitive searches Downloading and integration of patent legal status Downloading and integration of graphics Downloading and integration of pdf documents
  73. 73. Software for Technological Patent Intelligence- 74 - 2.- Filtering andValue Adding Automatic duplicate detection and removal Automatic grouping of patent families Automatic generation of field indexes Ability to define and build new indexes Wizard for grouping and cleaning terms of indexes Patent pertinence (user filled field) Annotation of patents (user filled field) Ability to define and edit patent groups Links to other related documents Taxonomies creation and edition 3.- Local Analysis and Exploitation Automatic extraction of main keywords from patents Automatic abstracts Automatic clustering of patents Automatic classification of patents using semantic filters Full text searching capabilities Semantic searching capabilities 4.- Graphic Generation Cite Analysis (cited and citing patents in relation to a known patent) Rankings - Analysis of one field. Matrix or Bar graphs – Two field’s co- occurrence analysis. Network relations analysis – Two fields co- occurrence analysis Space or topographic representation of a patent collection – text mining analysis
  74. 74. - 75 -Comparison of Software: Supply - Matheo Patent v7.1 Ability to use local databases to integrate new data and complete the patent analysis 5.- Dissemination and Workgroup Ability to publish the contents in the intranet / internet Personalised alerts Alerts to detect changes in the legal status of a patent Automatic reports using templates Ability to export data Ability to create a poll and link a patent to a poll Ability to link a patent to a forum Ability to link a patent to an event in a shared agenda 6.- Management ofTool Management of users access rights Management of Document collections access rights Simultaneous multi-user access and edition Customization of access and search interface Multilanguage interface System utilization statistics Table 13 – “Matheo Patent v7.1” Benchmark. 5.2.1. Definition of the software Matheo Patent is a tool specifically designed for Technological Supervision and Management of Industrial Property, which allows the automization of the use of various sources of patents (Espacenet, USPTO requests and USPTO concessions). In addition, it manages the publishing, annotation, grouping
  75. 75. Software for Technological Patent Intelligence- 76 - and consultation of patents found, as well as updating them. Finally, it is capable of analysing patents from a variety of viewpoints, generating matrices, histograms or relationship mapping. Figure 19 – “Matheo Patent v7.1” main screen. 5.2.2. Comments on the features studied 1) Searching and Downloading CDatabase connections: Matheo Patent contains a module/interface to carry out searches in Espacenet and USPTO. Matheo Patent breaks down the searches into time periods: year to year or even month to month. Thus, even with wide search strategies, it does not go beyond the visualisation limit of 500 patents imposed by Espacenet. It is therefore capable of retrieving all existing information. It can also download all the fields available for each patent (bibliographic file summary, search requests, graphics, first page and even a PDF file with the whole patent document). These fields can then be stored in a local database. Matheo Patent also allows the user to carry out various complementary search strategies in various phases given that, once a particular set of search results has been downloaded, a new question can be asked and new patents can be downloaded. Patents already downloaded from an earlier search will
  76. 76. - 77 -Comparison of Software: Supply - Matheo Patent v7.1 not be repeated. It also permits the selective loading of a specific list of patents of interest. At any given moment and for any patent, Matheo Patent permits the download of those fields which were not initially done so. Figure 20 – Search form for “Matheo Patent v7.1”. At any given moment, the user may carry out any of the above search strategies for the project and download the new patents published concerning that subject. Matheo Publisher detects the date on which this search was last run and runs the search only from that date onwards. All new patents are assigned an icon indicating that this information is pending revision. 2) Filtering and Value Adding Identifying duplicates: Matheo Patent detects duplicated patents and does not download them. Working with fields: Matheo Patent cannot create new fields based on existing ones. Neither can it create new indices. Matheo Patent has an internal engine allowing it to carry out advanced searches in all fields using Boolean Logic. This function facilitates the identification of patents fulfilling certain conditions and thus permits the creation of groups. Standardizing and cleaning indices: Matheo Patent automatically generates
  77. 77. Software for Technological Patent Intelligence- 78 - 9 indices (inventor, applicant, year of priority, year of publication, number of family members, group, 4-digit IPC classification, complete IPC classification and ECLA classification) making them available to users so they may check them and examine the patents classified within each term. Fields with information downloaded from the databases cannot be edited, except for ‘inventor’ and ‘applicant’. There is no help facility for editing inventors and applicants. If a user wishes to purge an index, the non-standard entries must be corrected manually. Figure 21 – “Matheo Patent v7.1” index. Thesaurus: Matheo Patent doe not permit the creation of lists of key words, a thesaurus or lists of empty words. Text-mining: Matheo Patent does not dispose of text-mining technology. The title, summary and queries can be loaded but may not be analysed. User classification: The user can give value to the information by adding notes to any patent.To do this, the “comments” field is used. It is also possible to evaluate a patent’s relevance (on a scale of 1 to 8) and create customized groups. Users cannot create a link between each patent and its pdf document format, although the patent’s first page can be linked with the “mosaic” page, which contains the most representative graphics and diagrams. 3) Local Analysis and Exploitation Matheo Patent includes the “Report” function which creates a new document
  78. 78. - 79 -Comparison of Software: Supply - Matheo Patent v7.1 in MS word, with a variety of options to choose from: - Quick Report: creates a document with the project title, the number of families, the number of patents, the search strategies used and a range of calculations for each of the most significant fields. - Global Report: this may include the following sections at the user’s discretion: General: overall data on the project. Details (histograms): Main inventors, applicants, 4-digit IPC classifications, 7-digit CPIs, complete Consumer Price Indices and ECLA classifications. Statistics (main co-occurrences): Inventors/applicants, inventors/4-digit Consumer Price Indices (from now on, CPIs), applicants/4-digit IPCs. User information: Histogram with the number of patents per group and lists of the number of patents included in each group. - IPC Report: Includes analysis focused on the IPC field: General: Overall project data. Details (histograms): 4-digit CPI, 7-digit CPI, complete CPI. Statistics (main matches): 4-digit CPI/Applicant, 4-digit CPI/ Year of publication, 7-digit CPI/Applicant, 7-digit CPI/Year of publication, 7-digit CPI/7-digit CPI, complete CPI/ complete CPI. Matrices (tables): 4-digit CPI/Applicant, 4-digit CPI/Year of publication, 7-digit CPI/Applicant, 7-digit CPI/Year of publication. - List: There are four kinds of lists: Inventors: list of inventors with their frequency. IPC Class 4 digits: lists the 4-digit CPI classifications and their frequency. IPC Class all digits: lists the 4-digit CPI classifications and their frequency. Patent assignee: lists the patentees and their frequency. - Patent Assignee Report: includes analysis focused on the applicant field: General: Overall project data. Details: Histogram with main applicants and lists of all applicants by frequency. Statistics (main matches): Applicants/7-digit CPIs, applicants/
  79. 79. Software for Technological Patent Intelligence- 80 - complete CPIs, applicants/year of publication. - Short Report: Includes the following analysis: General: Overall project data. Statistics (main matches): Applicants/year, applicants/4- digit CPIs, applicants/complete CPIs, 4-digit CPIs/year of publication, complete CPI/year of publication. 4) Graphic Generation Matheo Patent can create the following types of graphs: Chart (histogram): Corresponds to the analysis of a field’s content. The vertical axis. Represents the number of patents which correspond to each term. The graph can easily be set to limit the minimum frequency of each term. Figure 22 - “Matheo Patent v7.1” chart. Matriice (Matrix): Carries out analysis of matches in two fields. The fields which can be used for this analysis are: inventor, applicant, year of priority, year of publication, number of family members, group, 4-diit CPI classification, complete IPC classification and ECLA classification. The result is a matrix of cells in which the match number appears for each one. The cells’ colour becomes more intense the higher the number of matches.
  80. 80. - 81 -Comparison of Software: Supply - Matheo Patent v7.1 Figure 23 – The “Matheo Patent v7.1” matrix. Network: This is a 2-dimensional map diagram of the main elements in two fields and the relationship between them. - Each element is shown as a coloured rectangle with its name. - Each element has a number indicating its frequency (the number of patents in which it appears). Every element can have a link to another one in another field. In this case, they form a “pair”. This pair has a reference number indicating its co-occurrence (the number of patents in which the two terms appear). The user can create a network from one field (for instance, to analyse the relationship between inventors) or from two. In the latter case, the relationship between companies and the groups of patents they have created are shown.
  81. 81. Software for Technological Patent Intelligence- 82 - Figure 24 – “Matheo Patent v7.1” network. 5) Dissemination and Workgroup Matheo Patent is designed for personal use and does not have at its disposal functions enabling various users to interact. Neither is it possible to publish information through the web in order for others to perform a search for that information from other terminals. Matheo Patent can, however, export any group of patents to MS Word, indicating the fields which a user wants to include. Bibliographical references can also be exported in text format (indicating the information about each field with a label) or in .xml format, so that they can be incorporated into another application on an intranet. 6) Management of Tool The functions mentioned in this section refer to software accessed by various users at once. Matheo Patent does not have these functions.
  82. 82. - 83 -Comparison of Software: Supply - PatentLabII v1.41 5.3 Program evaluated: PatentLabII v1.41 Producer: Wisdomain Inc. 2300 North Barrington Road, Suite 400 Hoffman Estates, IL 60195 Tel: 1.847.490.5310 Fax: 1.847.885.7965 Website: www.wisdomain.com PATENTLAB II v1.41.0 + LaaMerger + LabViewer Evaluation 1 2 3 4 5 1.- Searching and Downloading Ability to search in a set of online patent databases Ability to search in other technical/grey literature online databases Ability to search in local (intranet) databases Ability to import patent records Ability to import other records (not patents) Ability to launch simultaneous searches in multiple databases Ability to save search strategies Ability to Schedule repetitive searches Downloading and integration of patent legal status Downloading and integration of graphics Downloading and integration of pdf documents
  83. 83. Software for Technological Patent Intelligence- 84 - 2.- Filtering andValue Adding Automatic duplicate detection and removal Automatic grouping of patent families Automatic generation of field indexes Ability to define and build new indexes Wizard for grouping and cleaning terms of indexes Patent pertinence (user filled field) Annotation of patents (user filled field) Ability to define and edit patent groups Links to other related documents Taxonomies creation and edition 3.- Local Analysis and Exploitation Automatic extraction of main keywords from patents Automatic abstracts Automatic clustering of patents Automatic classification of patents using semantic filters Full text searching capabilities Semantic searching capabilities 4.- Graphic Generation Cite Analysis (cited and citing patents in relation to a known patent) Rankings - Analysis of one field. Matrix or Bar graphs – Two field’s co- occurrence analysis Network relations analysis – Two fields co- occurrence analysis Space or topographic representation of a patent collection – text mining analysis
  84. 84. - 85 -Comparison of Software: Supply - PatentLabII v1.41 Ability to use local databases to integrate new data and complete the patent analysis 5.- Dissemination and Workgroup Ability to publish the contents in the intranet / internet Personalised alerts Alerts to detect changes in the legal status of a patent Automatic reports using templates Ability to export data Ability to create a poll and link a patent to a poll Ability to link a patent to a forum Ability to link a patent to an event in a shared agenda 6.- Management of Tool Management of users access rights Management of Document collections access rights Simultaneous multi-user access and edition Customization of access and search interface Multilanguage interface System utilization statistics Table 13 – “PatentLabIIv1.4l” Benchmark. 5.3.1. Definition of Software PatentLab is a tool used for the statistical analysis and visualisation of registry details from patents obtained through the Delphion patent search service. This site offers the following collections of patents: PCT, European (request & concession), North American (request and concession), German, Japanese
  85. 85. Software for Technological Patent Intelligence- 86 - and Inpadoc. 5.3.2. Comments on the features studied 1) Searching and Downloading Connection to databases: PatentLabII does not contain modules to connect and carry out searches in Delphion. When a search has been undertaken in Delphion, the results to be analysed have to be downloaded to a file (regardless of the number of registers) in “.laa” format in order to be used by this software. If the user wishes to combine several “.laa” files obtained through various searches and then carry out an analysis on that group of patents, they need to use the LaaMerger/LabViewer program, allowing the merger of two “.laa” files. Figure 25 – Laa Merger in “PatentLabIIv1.41”. PPatentLab cannot connect to Delphion to check or complete the information. PatentLabII can load the majority of fields with text information from Delphion but is unable to load the graphics connected to each registry. 2) Filtering and Value Adding Identification of duplicates: LaaMerger has a system for detecting duplicated patents from several “.laa” files. PatentLabII does not provide any system for detecting duplicated patents, nor for grouping patents from the same family in
  86. 86. - 87 -Comparison of Software: Supply - PatentLabII v1.41 a single registry. These operations should be undertaken before downloading the collection of patents. Working with fields: PatentLabII cannot create new fields from existing ones, nor can it create new indices. The majority of the information in each registry can be edited. PatenLabIIv1.41 does not permit the search for and consequent editing of particular registers. It only allows searches which are part of the filtering process taking place before the analysis is carried out. Standardization and cleaning of indices: PatentLabII has a feature which allows the detection of patentees with identical names. Furthermore, it permits the cutting and pasting of one name on top of another, considering this action as a “Non-normalized variant”. Figure 26 – “PatentLabIIv1.41” edit assignee. Thesaurus: PatentLabII does not allow the creation of lists of key words, thesauruses, empty words, etc. Text-mining: PatentLabII does not have text-mining technology. The title, summary and enquiries can be loaded in PatentLabII but cannot be analysed. User Classifications:There are four fields into which the user can download information of his choice. However, there is no tool to ease this task or to guarantee that the classifications are added without errors.
  87. 87. Software for Technological Patent Intelligence- 88 - Figure 27 – Editing a registry-1 in “PatentLabIIv1.41”. Figure 28 – Editing a registry-2 in “PatentLabIIv1.41”. PatentLabII does not have specific fields for making notes, nor for the evaluation of a patent’s importance. However, the “fields defined by the user” can be used for this purpose. It is also not possible to include a link for each patent to its corresponding .pdf document.
  88. 88. - 89 -Comparison of Software: Supply - PatentLabII v1.41 3) Local Analysis and Exploitation PatentLabII v.1.41 has an assistant to create the most common matrices and graphics. This assistant does not offer the full range of combinations of fields. Nevertheless, it is capable of creating matrices and graphics manually selecting any pair of fields. - The LabViewer facility allows the user to visualize all the registers from the database sequentially. It does not permit a search but allows the listing of the fields of most interest from a group of marked registers. - There is a “report” function with various standard options: - Overall Summary: Creates tables with an analysis of the following fields: patentee, Inventor, country, year and classification, showing the most significant ones. - Assignee Summary: Creates the following tables: patentee-year, patentee-country, Patentee-country-year, patentee-classification and patentee-main inventor. - Assignee Detail – Patent Classification: Creates the following tables: Main Classification-patentee, main associated classification-patentee and patentee-original classification-crossed classification. - Patent Classification Summary: Creates the following tables: Classification-year and Original classification-crossed classification. - Country Summary: Creates the following tables: Classification- country, country-year and country-patentee. 4) Graphic Generation PatentLabII can create graphics in two or three dimensions. The graphics in two dimensions correspond with the statistical analysis of a field’s content in which the height of each bar is proportional to the number of patents corresponding to each term. By clicking twice on each element, the list of patents corresponding to that term appears. The three-dimensional graphics correspond to analysis of the co- occurrences between two fields. The height of the bar is proportional to the number of co-occurrences existing between each pair of terms. By clicking twice on each cell, the list of patents corresponding to these terms appears.
  89. 89. Software for Technological Patent Intelligence- 90 - Figure 29 – 2-dimensional bar graph, “PatentLabIIv.1.41”. Figure 30 – 3-dimensional bar graph “PatentLabIIv.1.41”.
  90. 90. - 91 -Comparison of Software: Supply - PatentLabII v1.41 Thanks to the existence of personalized fields, PatentLabII allows the creation of totally personalized graphs and matrices. In this case, the following fields have been created: risk of infringement of a patent (high/medium/low) and the difficulty of a technology (high/medium/low), leading to the following examples below in figures 31 and 32. Figure 31 – Competitor – Risk of Infringement graph from “PatentLabIIv.1.41”. Figure 32 – Technological Difficulty – Risk of Infringement from “PatentLabIIv.1.41”.
  91. 91. Software for Technological Patent Intelligence- 92 - 5) Dissemination and Workgroup PatentLabII is designed for personal use and does not have functions enabling various users to interact. Neither is it possible to publish information via the web in order for others to visualize information from other terminals. PatentLabII can directly export areas of interest from matrices to MS Excel and can also generate documents in HTML format, which can then be published directly on an intranet. 6) Management of Tool The functions mentioned in this section are of relevance only to those programs designed for use by several people at once. PatentLabII does not offer these functions.
  92. 92. - 93 -Comparison of Software: Supply - PM Manager v1.4.0.3 5.4 Program Evaluated: PM Manager v1.4.0.3 Producer: WIPS Co. Ltd. 93-45 Bookchang-Dong, Joong-Gu, Seoul 100-080 Republic of Korea TEL: +82-(0)2-726-1103/1109 FAX: +82-(0)2-726-1001 Email: global@wips.co.kr Website : www.wipsglobal.com PM MANAGER v1.4.0.3 Evaluation 1 2 3 4 5 1.- Searching and Downloading Ability to search in a set of online patent databases Ability to search in other technical/grey literature online databases Ability to search in local (intranet) databases Ability to import patent records Ability to import other records (not patents) Ability to launch simultaneous searches in multiple databases Ability to save search strategies Ability to Schedule repetitive searches Downloading and integration of patent legal status Downloading and integration of graphics Downloading and integration of pdf documents
  93. 93. Software for Technological Patent Intelligence- 94 - 2.- Filtering andValue Adding Automatic duplicate detection and removal Automatic grouping of patent families Automatic generation of field indexes Ability to define and build new indexes Wizard for grouping and cleaning terms of indexes Patent pertinence (user filled field) Annotation of patents (user filled field) Ability to define and edit patent groups Links to other related documents Taxonomies creation and edition 3.- Local Analysis and Exploitation Automatic extraction of main keywords from patents Automatic abstracts Automatic clustering of patents Automatic classification of patents using semantic filters Full text searching capabilities Semantic searching capabilities 4.- Graphic Generation Cite Analysis (cited and citing patents in relation to a known patent) Rankings - Analysis of one field. Matrix or Bar graphs – Two field’s co- occurrence analysis. Network relations analysis – Two fields co- occurrence analysis Space or topographic representation of a patent collection – text mining analysis
  94. 94. - 95 -Comparison of Software: Supply - PM Manager v1.4.0.3 Ability to use local databases to integrate new data and complete the patent analysis 5.- Dissemination and Workgroup Ability to publish the contents in the intranet / internet Personalised alerts Alerts to detect changes in the legal status of a patent Automatic reports using templates Ability to export data Ability to create a poll and link a patent to a poll Ability to link a patent to a forum Ability to link a patent to an event in a shared agenda 6.- Management of Tool Management of users access rights Management of Document collections access rights Simultaneous multi-user access and edition Customization of access and search interface Multilanguage interface System utilization statistics Table 14 – “PM Manager v1.4.0.3” Benchmark. 5.4.1 Definition and positioning of software PM Manager has been developed as a complement to the WIPS Global patent search system. The present global reach of this system is: Patents in Korea, Japan, China, USA, Pat. Europe, Pat. PCT, Inpadoc and GPAT (Switzerland, France, Great Britain and Germany).
  95. 95. Software for Technological Patent Intelligence- 96 - PM Manager is able to load files obtained from WIPS Global and then, classify, analyse and process information from patents by applying different viewpoints. PM Manager’s focus is on giving the user the capacity to modify, annotate and complete information by adding new personalized fields. It then allows basic statistical analysis, advanced statistical analysis and other analysis tailor-made for the personalized fields created by the user. 5.4.2 Comments on the features studied 1) Searching and Downloading Connection to databases: PM Manager does not dispose of modules to connect and run search in WIPS Global. Its starting point is the importation of a list of already existing registers in “.pmd” format, downloaded from WIPS Global. Once a search has been run on WIPS Global the results can be downloaded (in groups of 200 registers at the most) in “.pmd” format to be treated by the program. PM Manager is equipped to connect with WIPS Global to check for new information (e.g. whether a family of patents has changed). Figure 33 – Main screen, “PM Manager v1.4.0.3”.
  96. 96. - 97 -Comparison of Software: Supply - PM Manager v1.4.0.3 Importation: PM Manager can also import registers in electronic spreadsheet format (.xls) PM Manager is able to load an associated graph in each register. If this information is not available, it can connect to WIPS Global to check whether a graph is available and if so, download it. 2) Filtering and Value Adding Identification of duplicates: New lists of patents in “.pmd” format may be added (merged). If there are equivalent patents, the system will detect them automatically and ask if the user wishes to keep the existing version, substitute it for the new one or complete the present fields with the new information obtained. Families of patents: PM Manager can identify members of the same family through the priority number. Working with fields: PM Manager cannot create new fields from existing ones, nor can it create new indices. Nevertheless, all information loaded can be edited. Standardization and cleaning of indices: PM Manager has a feature allowing the selection of the names of “non-unified” patentee companies which are, in the user’s judgement, equivalent. The user can then key in the name “unified”, which is then saved for future use. It is also possible to import lists of unified companies from previous projects. Thus, the list grows as time passes. Figure 34 – Index cleaner, “PM Manager v1.4.0.3”.
  97. 97. Software for Technological Patent Intelligence- 98 - Thesaurus: PM Manager does not permit the creation of lists of key words, thesaurus or empty words. Text mining: PM Manager does not have text mining technology. User Classifications: The software allows for the creation of a three-level thematic classification which can be as extensive as required. In order to facilitate this task, this structure appears when the user wishes to classify the content of a patent. PM Manager also permits the user to define up to 5 kinds of classifications to evaluate a patent, together with its variables. These classifications can be whatever the user wishes. As an example, the following can be cited: - possibility of infraction (high/medium/low). - price of technology (expensive/mid/low). - difficulty of technology (complex/normal/simple). Figure 35 – User classification, “PM Manager v1.4.0.3”. PM Manager also contains another field to evaluate the importance of the patent (A/B/C/D), a “memo” field in which notes can be included and a “core patent” field, used to label a patent as “key”.
  98. 98. - 99 -Comparison of Software: Supply - PM Manager v1.4.0.3 Last but not least, with each patent a range of links with documents and applications can be included. Figure 36 – Edit key information in “PM Manager v1.4.0.3”. 3) Local Analysis and Exploitation PM Manager offers several functions to carry out all kinds of analysis: - The “Create Technology development Map” function highlights one or several of the technological classifications of interest and indicates to the user how these technologies have evolved over time, showing for each year the number of patents, the titles and the patentees of the technologies (see Figure 37). - The “Create Key Information List” function automatically creates a report in MS Word format with all registries of patents marked “key” including the “purpose of the patent”, the main claim”, “problems with the state of the previous technique and the “explanation of the graph”. This report is of value when evaluating and comparing technologies competing among themselves (see Figure 38).
  99. 99. Software for Technological Patent Intelligence- 100 - Figure 37 – Technology Development Map, “PM Manager v1.4.0.3”. Figure 38 – Key Information List, “PM Manager V1.4.0.3”. 4) Graphic Generation MP Manager can create graphs in two or three dimensions. The two- dimensional graphs correspond to the statistical analysis of a field’s content, in

×