Concept-based, semantic searchAndreas BlumauerSemantic Web Companywww.semantic-web.at                  © Semantic Web Comp...
Content/agenda1. What means „concept-based“?2. Concept-tagging3. Semantic search  •   Faceted search  •   Similarity searc...
What is a concept?                The semiotic triangle                                                                   ...
Concept-based enterprise vocabulary                           http://voc.org.com/core/355           http://voc.org.com/cor...
Concept-tagging vs. Term-taggingConcept-tagging is done on top             Enterprise vocabularyof concepts which are alre...
Concept-tagging: pre-condition for                        semantic search                                                 ...
Traditional search methods vs.                                    semantic search                                    W 176...
Semantics as a means for                                     interpretationSemantics helps to makedifferent language level...
Concept-based high-precision facet                               classification#1     ---- --- -- --      Daimler-Benz ---...
Similarity search: efficient re-use of                                existing information                   Mercedes-AMG ...
Topic Pages: Mashups for a                 fast 360O view                                                                 ...
Linked Data: complex queries on top of                      standard technologiesExample: Find industry news which mention...
Conclusio 1: The three levels of                                 semantic searchYear in which the    2014                 ...
Conclusio 2: Explicit metadata layer                          Data                                  Data    Research      ...
“Thank you for your time and                                           please forward any comments                        ...
Upcoming SlideShare
Loading in...5
×

Concept based semantic search

1,132

Published on

1 Comment
3 Likes
Statistics
Notes
  • Thank you for nice and intersting presentaion.
    I mean that at first any information source shloud be linked to class (or domain name). Then it should be linked to different terms (concepts) and their values. I mean that we should differ concepts as terms and conceptes as entities (names and values).. Concepts and entities names should be linked by different relationships (associations). Entity values as well. I am collecting a lot of information using Topic map model and try to uderstan all details.
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here
No Downloads
Views
Total Views
1,132
On Slideshare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
34
Comments
1
Likes
3
Embeds 0
No embeds

No notes for slide

Concept based semantic search

  1. 1. Concept-based, semantic searchAndreas BlumauerSemantic Web Companywww.semantic-web.at © Semantic Web Company – http://www.semantic-web.at/ 1
  2. 2. Content/agenda1. What means „concept-based“?2. Concept-tagging3. Semantic search • Faceted search • Similarity search4. Semantics as a means for ‚interpretation‘5. Topic pages6. Three levels of semantic search © Semantic Web Company – http://www.semantic-web.at/ 2
  3. 3. What is a concept? The semiotic triangle Mental model of „A-Class“ concept Another mental model of „A-Class“another objectA-ClassA-Klasse label object W 176 © Semantic Web Company – http://www.semantic-web.at/ 3
  4. 4. Concept-based enterprise vocabulary http://voc.org.com/core/355 http://voc.org.com/core/54 Vehicle prefLabel prefLabel manufacturing compact car company broader broader Daimler-Benz A-Class prefLabel related prefLabel (de) Daimler AG A-Klasse http://voc.org.com/core/97 http://voc.org.com/core/176 W 176 narrower narrower Mercedes-AMG prefLabel related prefLabel AMG A 250 Sport http://voc.org.com/core/77 http://voc.org.com/core/44Each concept has a unique URI and can have various multi-lingual labels. Additionaly, it can have various types ofsemantic relations with other concepts. 4 W3C´s SKOS standard describes a pre-defined set of semantic relations especiallyfor controlled vocabularies. © Semantic Web Company – http://www.semantic-web.at/ 4
  5. 5. Concept-tagging vs. Term-taggingConcept-tagging is done on top Enterprise vocabularyof concepts which are alreadypart of the enterprisevocabulary, thus contextualised ‚Term-tags„ become a ‚concept„and linked to other concepts. as part of the enterprise vocabularyTerm-tagging means that tagsare extracted from text(automatically via text mining)which are not part of the Concept Taggingcontrolled vocabulary yet. --- ------ - Term TaggingTerm-tags can be inserted intothe enterprise vocabulary. -- --- ---- -This extends and refines the ---- ---- ---vocabulary more and more. ---- --- - -- - --- ---- -- --- ------ Content from CMS © Semantic Web Company – http://www.semantic-web.at/ 5
  6. 6. Concept-tagging: pre-condition for semantic search W 176 search--- -- ----- -- prefLabel------ ---- --- A-Class------ --- ---- -- --- -- narrower W 176A 250 Sport ---- - ---- ---- ---- ---- --- prefLabel A 250 Sport © Semantic Web Company – http://www.semantic-web.at/ 6
  7. 7. Traditional search methods vs. semantic search W 176 search Semantic: prefLabel Can the search phrase A-Class be found analogously?Traditional: narrowerCan the search phrase W 176be found literallyin the document? prefLabel A 250 Sport --- -- ----- -- --- -- -- --- - ------ ---- --- ----- --- ----- ------ --- ---- - --- ---- --- -- --- --- ---- ----------A 250 A 250 Sport ---- - Sport ---- ----- ---- ---- ---- ---- ---- ---- ---- --- --- © Semantic Web Company – http://www.semantic-web.at/ 7
  8. 8. Semantics as a means for interpretationSemantics helps to makedifferent language levels or W 176 searchvarious perspectivescomparable. prefLabel A-ClassExample: Vendors and theircustomers quite often talk narrower W 176different languages. Wrong orsometimes time-consuming‚translations„ and prefLabel A 250 Sportinterpretations have to be doneby the customers themselves.Example: The state of ----- --- -----knowledge of employees can be - --- ---- ----quite divergent. Semantics as a --- --- -A 250search assistant can serveespecially less experienced Sport ---- -----colleagues. ---- ---- ---- --- © Semantic Web Company – http://www.semantic-web.at/ 8
  9. 9. Concept-based high-precision facet classification#1 ---- --- -- -- Daimler-Benz ----- Synonyms and hidden labels: #1 is also classified as ‚Daimler - --- ------ --- AG„ because ‚Daimler-Benz„ is also (an old) name for ‚Daimler - ----- ---- --- AG„. - ---- ------ -- Transitivity: COMPANY #2 is categorized as ‚vehicle manufacturer„ too, because in#2 ----- ------ -- our thesaurus ‚AMG„ is narrower Vehicle manufacturer (2) (is part of) of ‚Daimler„ which is a - ------ -- --- ‚vehicle manufacturer„. ---- ---- ----- Daimler AG (2) ---- ---- ---- AMG -- AMG (1) ---- --- ------ --Concept-/thesaurus-based facet classification of documents is as precise as the classification schemeused by the enterprise thesaurus itself. In consideration of all different labels of concepts and theirtransitive hierarchical relations, a more precise facet classification can be realised than withtraditional term-based methods. 9 © Semantic Web Company – http://www.semantic-web.at/ 9
  10. 10. Similarity search: efficient re-use of existing information Mercedes-AMG --- -- AMG http://voc.org.com/core/77 --- ------ --- prefLabel ------ -- ---- AMG -- ---- ----- - -- --A 250 Sport - --- ----- ---- http://voc.org.com/core/176 --- ---- ----- -- --- --- -- --- ------ -- A-Class ---- ---- --- - Mercedes-AMG -------- ----- -------- --- -------- -- W 176 W 176 ---- narrower ---- ----- ---- ---- ---- --- A 250 Sport http://voc.org.com/core/44Content-authors as well as end-users can benefit from similarity search (content recommendation),e.g. by ‚skim reading„ or by the avoidance of duplicated work. Even if two documents have no words incommon they can be classified as similar when using a concept-based text analysis. 10 © Semantic Web Company – http://www.semantic-web.at/ 10
  11. 11. Topic Pages: Mashups for a fast 360O view Articles (twitter, videos etc.) can be retrieved Short http:/ from various content sourcesdescription / Related concepts CMSGeo search API 11 © Semantic Web Company – http://www.semantic-web.at/ 11
  12. 12. Linked Data: complex queries on top of standard technologiesExample: Find industry news which mention countries or regions, in which our exportvolume increased by more than 10% over the last 5 years an which mention either one ofour products and/or a competitor. (Federated) SPARQL Queries Industry Export statistics News 12 © Semantic Web Company – http://www.semantic-web.at/ 12
  13. 13. Conclusio 1: The three levels of semantic searchYear in which the 2014 Semantics is explicitly available via linked knowledge models.underlying Content from various sources and deparments can be linked andtechnology will Linked Data mashed on top of an explicit meta data layer. Complex queriesbe/has been rolled based search which use data from many sources can be made by using theout. standard query language SPARQL. 2011 Semantics is explicitly available by using controlled vocabularies and thesauri. Thesauri are the basis for precise text analysis and Concept- to build a semantic index. Building knowledge models is based search especially cost-efficient for larger organisations since a more precise search can be provided. No Standards 2005 Semantics is calculated by text analysis. Example: Because Term-based „Dieter Zetsche“ frequently occurs together with „Daimler AG“ in a text the algorithm assumes that those two phrases relate search somehow to each other. Term-based methods are less precise than the two from further above. © Semantic Web Company – http://www.semantic-web.at/ 13
  14. 14. Conclusio 2: Explicit metadata layer Data Data Research Production Metadata:• Stored and processed separately from data• Metadata management is part of the enterprise information management strategy Data Data Marketing/Sales HR © Semantic Web Company – http://www.semantic-web.at/ 14
  15. 15. “Thank you for your time and please forward any comments or questions to me to get more information on our product or linked data & vocabularies!”Andreas BlumauerManaging Partnera.blumauer@semantic-web.atSemantic Web Company GmbH http://www.semantic-web.at/Mariahilfer Strasse 70/8 http://poolparty.biz1070 ViennaAustria http://twitter.com/semwebcompany © Semantic Web Company – http://www.semantic-web.at/ 15 15
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×