IBM WATSON™
NATURAL LANGUAGE UNDERSTANDING
+
MIND MAPPING
NLU-MAP
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 1
Image courtesy of hyward at
FreeDigitalPhotos.net
IBM WATSON™
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 2
Capabilities not found in traditional computing
systems
• Understand like humans do, processing
natural language and other unstructured data.
• Learn, getting more valuable with time.
• Reason. It understands underlying ideas and
concepts, form hypothesis, infers and extracts
concepts.
• Interact. It has abilities to see, talk and hear. It
can interact with humans in a natural way.
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 3
What can you do with IBM Watson™?
• Analyze and interpret all of your data,
including unstructured text, images, audio and
video.
• Use machine learning to grow the expertise in
your apps and systems.
• Provide personalized recommendations by
understanding the user’s personality, tone,
and emotion
• Create chat bots that can engage in dialog.
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 4
IBM WATSON™ NLU
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 5
Natural Language Understanding (NLU)
SEMANTIC ANALYSIS OF TEXT TO:
• Get insight
• Understand sentiment and emotion
Watson™ NLU Overview
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 6
Sources of information to analyze
• Social media & Blogs
• Articles
• Research reports
• Enterprise mail and e-mail
• Surveys
• Documents
• Voice transcriptions
• Chat
• News
• Knowledge bases
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 7
Types of use
• Social media monitoring
• Content recommendation
• Opinion mining
• Content profiling
• Add placement
• Buyer intent analysis
• Churn prevention
• Financial prediction
• Brand & product intelligence
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 8
Results of the analysis
• Entities
• Relations
• Keywords
• Concepts
• Categories
• Semantic roles
• Metadata
• Sentiment
• Emotion
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 9
Problems
• Too much information to visualize as linear
text.
• Disorientation when visualizing the results as
web pages.
• Lack of the whole picture view in both cases.
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 10
Solution
IBM Watson™ Natural Language Understanding
+ Mind Mapping automation
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 11
Introduction to mind mapping
NLU-MAP
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 12
NLU-MAP
SaaS product developed by Infoseg using IBM Watson™ NLU API
and our own software to automate the creation of mind maps.
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 13
Sample text to analyze
A Google takeover of at least part of HTC's business has been rumoured for some time.
Now that the Taiwanese firm has announced its shares are being suspended, speculation is rampant that an agreement has
been struck.
HTC made the first ever Android handset - the Dream - and is rumoured to be the manufacturer of one of the US firm's Pixel 2
models, which is set to be announced next month.
But Google has already struggled to integrate one phone-maker, Motorola Mobility, and it's not clear why it would want to
repeat the experience.
Yes, HTC has proven itself capable of developing unusual features - such as the squeeze-to-take-photos design of its recent
U11 - but it has repeatedly failed to launch a bestseller. And does Google really want to own HTC's factories at a time when
others, including Apple, are happy to outsource production?
An alternative deal could involve buying HTC's virtual reality business.
Image copyright Getty Images Image caption Might a takeover deal be limited to HTC's virtual reality division?
Its high-end Vive headset is reportedly outselling Facebook's Oculus Rift rival by a margin of nearly two-to-one - albeit with
still modest numbers - and is recognised by many as the superior system.
Moreover, Google already poached Vive's chief designer Claude Zellweger away at the start of the year for its own Daydream
VR effort.
Many believe virtual and augmented reality (in which graphics are mixed with real-world views) have the potential to
revolutionise how we interact with computers, but only after further huge sums are invested in R&D.
Cash-strapped HTC might thus be incapable of making the most of its early lead, while Google may be willing to dig deep to
give itself every advantage possible against Microsoft and Apple, which are also eyeing the space.
http://www.bbc.com/news/technology-41336090
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 14
Mind map created by NLU-MAP
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 15
Text included as a Note
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 16
Categories
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 17
Concepts
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 18
Detail of Concepts
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 19
Entities
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 20
Sentiment
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 21
Relations
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 22
Keywords
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 23
Semantic Roles
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 24
FUTURE DEVELOPMENTS
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 25
Services provided by IBM Watson™
• Discovery. To unlock actionable insights hidden in
unstructured data. Also Natural Language
Understanding to extract semantic information
from content.
• Conversation and vision. Conversation, Chatbots,
Visual Recognition.
• Speech and Empathy. Speech to Text, Text to
Speech, Personality Insights, Tone Analyzer.
• Language. Translator, Natural Language Classifier,
Retrieve and Rank the most relevant information
from a collection of documents.
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 26
Next developments
• Tone Analyzer. This will be an add-in for Outlook
so that users can see the tone of the messages
they are going to send. In this way users will be
confident that their messages do not contain
improper content.
• Chatbots. To create a summary of conversations
that can be analyzed later.
• Personality Insights. To obtain a psychological
portrait of a person as a mind map.
• Discovery. To visualize the result of searches in
collections of documents.
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 27
Interview
Interview for the Mindmapping software blog
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 28
MIND MAPPING BOOKS WRITTEN
BY JOSE M. GUERRERO
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 29
Introduction to the Applications of Mind Mapping in Medicine
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 30
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 31
Introducción a la Técnica de Mapas Mentales
Gestión Visual de Información Compleja con
MindManager 16
http://www.editorialuoc.cat/introduccion-a-la-tecnica-de-mapas-mentales
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 32
Co-author of
Federal Data Science and Advanced Analytics in
Agricultural Science: Transforming Government
and Policy using Artificial Intelligence
Chapter 8
Elsevier – Academic Press
Infoseg
http://www.slideshare.net/jmgf2009/presentations
https://twitter.com/InfosegS
http://paper.li/InfosegS/1356259200
José M. Guerrero
jm@infoseg.com
https://es.linkedin.com/in/josemguerrero2012
C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 33

NLU-MAP. IBM Watson NLU with Mind Mapping automation

  • 1.
    IBM WATSON™ NATURAL LANGUAGEUNDERSTANDING + MIND MAPPING NLU-MAP C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 1 Image courtesy of hyward at FreeDigitalPhotos.net
  • 2.
    IBM WATSON™ C) Infoseg,S.A. 2017 http://bit.ly/1Eimh3k 2
  • 3.
    Capabilities not foundin traditional computing systems • Understand like humans do, processing natural language and other unstructured data. • Learn, getting more valuable with time. • Reason. It understands underlying ideas and concepts, form hypothesis, infers and extracts concepts. • Interact. It has abilities to see, talk and hear. It can interact with humans in a natural way. C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 3
  • 4.
    What can youdo with IBM Watson™? • Analyze and interpret all of your data, including unstructured text, images, audio and video. • Use machine learning to grow the expertise in your apps and systems. • Provide personalized recommendations by understanding the user’s personality, tone, and emotion • Create chat bots that can engage in dialog. C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 4
  • 5.
    IBM WATSON™ NLU C)Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 5
  • 6.
    Natural Language Understanding(NLU) SEMANTIC ANALYSIS OF TEXT TO: • Get insight • Understand sentiment and emotion Watson™ NLU Overview C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 6
  • 7.
    Sources of informationto analyze • Social media & Blogs • Articles • Research reports • Enterprise mail and e-mail • Surveys • Documents • Voice transcriptions • Chat • News • Knowledge bases C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 7
  • 8.
    Types of use •Social media monitoring • Content recommendation • Opinion mining • Content profiling • Add placement • Buyer intent analysis • Churn prevention • Financial prediction • Brand & product intelligence C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 8
  • 9.
    Results of theanalysis • Entities • Relations • Keywords • Concepts • Categories • Semantic roles • Metadata • Sentiment • Emotion C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 9
  • 10.
    Problems • Too muchinformation to visualize as linear text. • Disorientation when visualizing the results as web pages. • Lack of the whole picture view in both cases. C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 10
  • 11.
    Solution IBM Watson™ NaturalLanguage Understanding + Mind Mapping automation C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 11 Introduction to mind mapping
  • 12.
    NLU-MAP C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 12
  • 13.
    NLU-MAP SaaS product developedby Infoseg using IBM Watson™ NLU API and our own software to automate the creation of mind maps. C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 13
  • 14.
    Sample text toanalyze A Google takeover of at least part of HTC's business has been rumoured for some time. Now that the Taiwanese firm has announced its shares are being suspended, speculation is rampant that an agreement has been struck. HTC made the first ever Android handset - the Dream - and is rumoured to be the manufacturer of one of the US firm's Pixel 2 models, which is set to be announced next month. But Google has already struggled to integrate one phone-maker, Motorola Mobility, and it's not clear why it would want to repeat the experience. Yes, HTC has proven itself capable of developing unusual features - such as the squeeze-to-take-photos design of its recent U11 - but it has repeatedly failed to launch a bestseller. And does Google really want to own HTC's factories at a time when others, including Apple, are happy to outsource production? An alternative deal could involve buying HTC's virtual reality business. Image copyright Getty Images Image caption Might a takeover deal be limited to HTC's virtual reality division? Its high-end Vive headset is reportedly outselling Facebook's Oculus Rift rival by a margin of nearly two-to-one - albeit with still modest numbers - and is recognised by many as the superior system. Moreover, Google already poached Vive's chief designer Claude Zellweger away at the start of the year for its own Daydream VR effort. Many believe virtual and augmented reality (in which graphics are mixed with real-world views) have the potential to revolutionise how we interact with computers, but only after further huge sums are invested in R&D. Cash-strapped HTC might thus be incapable of making the most of its early lead, while Google may be willing to dig deep to give itself every advantage possible against Microsoft and Apple, which are also eyeing the space. http://www.bbc.com/news/technology-41336090 C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 14
  • 15.
    Mind map createdby NLU-MAP C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 15
  • 16.
    Text included asa Note C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 16
  • 17.
    Categories C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 17
  • 18.
    Concepts C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 18
  • 19.
    Detail of Concepts C)Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 19
  • 20.
    Entities C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 20
  • 21.
    Sentiment C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 21
  • 22.
    Relations C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 22
  • 23.
    Keywords C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 23
  • 24.
    Semantic Roles C) Infoseg,S.A. 2017 http://bit.ly/1Eimh3k 24
  • 25.
    FUTURE DEVELOPMENTS C) Infoseg,S.A. 2017 http://bit.ly/1Eimh3k 25
  • 26.
    Services provided byIBM Watson™ • Discovery. To unlock actionable insights hidden in unstructured data. Also Natural Language Understanding to extract semantic information from content. • Conversation and vision. Conversation, Chatbots, Visual Recognition. • Speech and Empathy. Speech to Text, Text to Speech, Personality Insights, Tone Analyzer. • Language. Translator, Natural Language Classifier, Retrieve and Rank the most relevant information from a collection of documents. C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 26
  • 27.
    Next developments • ToneAnalyzer. This will be an add-in for Outlook so that users can see the tone of the messages they are going to send. In this way users will be confident that their messages do not contain improper content. • Chatbots. To create a summary of conversations that can be analyzed later. • Personality Insights. To obtain a psychological portrait of a person as a mind map. • Discovery. To visualize the result of searches in collections of documents. C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 27
  • 28.
    Interview Interview for theMindmapping software blog C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 28
  • 29.
    MIND MAPPING BOOKSWRITTEN BY JOSE M. GUERRERO C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 29
  • 30.
    Introduction to theApplications of Mind Mapping in Medicine C) Infoseg, S.A. 2017 http://bit.ly/1Eimh3k 30
  • 31.
    C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 31 Introducción a la Técnica de Mapas Mentales Gestión Visual de Información Compleja con MindManager 16 http://www.editorialuoc.cat/introduccion-a-la-tecnica-de-mapas-mentales
  • 32.
    C) Infoseg, S.A.2017 http://bit.ly/1Eimh3k 32 Co-author of Federal Data Science and Advanced Analytics in Agricultural Science: Transforming Government and Policy using Artificial Intelligence Chapter 8 Elsevier – Academic Press
  • 33.