![Page 1: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/1.jpg)
Knowledge Graphs: In Theory and Practice
Sumit Bhatia1 and Nitish Aggarwal2
1 IBM Research, New Delhi, India2 IBM Watson, San Jose, CA
[email protected], [email protected] 10, 2017
CIKM 2017 Knowledge Graphs: In Theory and Practice 1/47
![Page 2: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/2.jpg)
Knowledge Graphs Analytics
![Page 3: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/3.jpg)
Knowledge Graph Analytics
• Finding Entities of Interest• Entity Search and Recommendation• Entity Linking and Disambiguation
• Entity exploration: Knowing more about the entities• Relationship Search• Path Ranking
• Upcoming challenges
CIKM 2017 Knowledge Graphs: In Theory and Practice 2/47
![Page 4: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/4.jpg)
Finding the Right Entities
CIKM 2017 Knowledge Graphs: In Theory and Practice 3/47
![Page 5: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/5.jpg)
Finding the Right Entities
CIKM 2017 Knowledge Graphs: In Theory and Practice 4/47
![Page 6: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/6.jpg)
Finding Right Entities
Entities are the fundamental units of a Knowledge graph. Howto get to the right entities in the graph?
Given a Knowledge Base, K = {E ,R}, a document corpus D,and a named entity mention m, map/link the mention m to itscorresponding entity e ∈ E .
SteveJobs
Apple
iPhone
PaloAlto
SteveWozniak
SteveBalmer
Seattle
BillGates
Windows
Microsoft
USA
Web Queries:steve jobs birthday
NL Questions:When did Steve resign fromMicrosoft?
NL Text:....Jobs and Wozniak started AppleComputers from their garage...
CIKM 2017 Knowledge Graphs: In Theory and Practice 5/47
![Page 7: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/7.jpg)
Finding Right Entities
Entities are the fundamental units of a Knowledge graph. Howto get to the right entities in the graph?Given a Knowledge Base, K = {E ,R}, a document corpus D,and a named entity mention m, map/link the mention m to itscorresponding entity e ∈ E .
SteveJobs
Apple
iPhone
PaloAlto
SteveWozniak
SteveBalmer
Seattle
BillGates
Windows
Microsoft
USA
Web Queries:steve jobs birthday
NL Questions:When did Steve resign fromMicrosoft?
NL Text:....Jobs and Wozniak started AppleComputers from their garage...
CIKM 2017 Knowledge Graphs: In Theory and Practice 5/47
![Page 8: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/8.jpg)
Finding Right Entities
Entities are the fundamental units of a Knowledge graph. Howto get to the right entities in the graph?Given a Knowledge Base, K = {E ,R}, a document corpus D,and a named entity mention m, map/link the mention m to itscorresponding entity e ∈ E .
SteveJobs
Apple
iPhone
PaloAlto
SteveWozniak
SteveBalmer
Seattle
BillGates
Windows
Microsoft
USA
Web Queries:steve jobs birthday
NL Questions:When did Steve resign fromMicrosoft?
NL Text:....Jobs and Wozniak started AppleComputers from their garage...
CIKM 2017 Knowledge Graphs: In Theory and Practice 5/47
![Page 9: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/9.jpg)
Finding Right Entities
Entities are the fundamental units of a Knowledge graph. Howto get to the right entities in the graph?Given a Knowledge Base, K = {E ,R}, a document corpus D,and a named entity mention m, map/link the mention m to itscorresponding entity e ∈ E .
SteveJobs
Apple
iPhone
PaloAlto
SteveWozniak
SteveBalmer
Seattle
BillGates
Windows
Microsoft
USA
Web Queries:steve jobs birthday
NL Questions:When did Steve resign fromMicrosoft?
NL Text:....Jobs and Wozniak started AppleComputers from their garage...
CIKM 2017 Knowledge Graphs: In Theory and Practice 5/47
![Page 10: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/10.jpg)
Finding Right Entities
Entities are the fundamental units of a Knowledge graph. Howto get to the right entities in the graph?Given a Knowledge Base, K = {E ,R}, a document corpus D,and a named entity mention m, map/link the mention m to itscorresponding entity e ∈ E .
SteveJobs
Apple
iPhone
PaloAlto
SteveWozniak
SteveBalmer
Seattle
BillGates
Windows
Microsoft
USA
Web Queries:steve jobs birthday
NL Questions:When did Steve resign fromMicrosoft?
NL Text:....Jobs and Wozniak started AppleComputers from their garage...
CIKM 2017 Knowledge Graphs: In Theory and Practice 5/47
![Page 11: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/11.jpg)
Finding Right Entities
Entities are the fundamental units of a Knowledge graph. Howto get to the right entities in the graph?Given a Knowledge Base, K = {E ,R}, a document corpus D,and a named entity mention m, map/link the mention m to itscorresponding entity e ∈ E .
SteveJobs
Apple
iPhone
PaloAlto
SteveWozniak
SteveBalmer
Seattle
BillGates
Windows
Microsoft
USA
Web Queries:steve jobs birthday
NL Questions:When did Steve resign fromMicrosoft?
NL Text:....Jobs and Wozniak started AppleComputers from their garage...
CIKM 2017 Knowledge Graphs: In Theory and Practice 5/47
![Page 12: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/12.jpg)
Challenges
• Same entity can be represented by multiple surface forms
Barack Obama, Barack H. Obama, President Obama,Senator ObamaPresident of the United States
• Same surface form could refer to multiple entitiesMichael Jordan – Basketball player or Berkeley professorwhen did steve leave apple?
• Out of KG mentions
CIKM 2017 Knowledge Graphs: In Theory and Practice 6/47
![Page 13: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/13.jpg)
Challenges
• Same entity can be represented by multiple surface formsBarack Obama, Barack H. Obama, President Obama,Senator Obama
President of the United States• Same surface form could refer to multiple entitiesMichael Jordan – Basketball player or Berkeley professorwhen did steve leave apple?
• Out of KG mentions
CIKM 2017 Knowledge Graphs: In Theory and Practice 6/47
![Page 14: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/14.jpg)
Challenges
• Same entity can be represented by multiple surface formsBarack Obama, Barack H. Obama, President Obama,Senator ObamaPresident of the United States
• Same surface form could refer to multiple entitiesMichael Jordan – Basketball player or Berkeley professorwhen did steve leave apple?
• Out of KG mentions
CIKM 2017 Knowledge Graphs: In Theory and Practice 6/47
![Page 15: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/15.jpg)
Challenges
• Same entity can be represented by multiple surface formsBarack Obama, Barack H. Obama, President Obama,Senator ObamaPresident of the United States
• Same surface form could refer to multiple entities
Michael Jordan – Basketball player or Berkeley professorwhen did steve leave apple?
• Out of KG mentions
CIKM 2017 Knowledge Graphs: In Theory and Practice 6/47
![Page 16: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/16.jpg)
Challenges
• Same entity can be represented by multiple surface formsBarack Obama, Barack H. Obama, President Obama,Senator ObamaPresident of the United States
• Same surface form could refer to multiple entitiesMichael Jordan – Basketball player or Berkeley professor
when did steve leave apple?• Out of KG mentions
CIKM 2017 Knowledge Graphs: In Theory and Practice 6/47
![Page 17: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/17.jpg)
Challenges
• Same entity can be represented by multiple surface formsBarack Obama, Barack H. Obama, President Obama,Senator ObamaPresident of the United States
• Same surface form could refer to multiple entitiesMichael Jordan – Basketball player or Berkeley professorwhen did steve leave apple?
• Out of KG mentions
CIKM 2017 Knowledge Graphs: In Theory and Practice 6/47
![Page 18: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/18.jpg)
Challenges
• Same entity can be represented by multiple surface formsBarack Obama, Barack H. Obama, President Obama,Senator ObamaPresident of the United States
• Same surface form could refer to multiple entitiesMichael Jordan – Basketball player or Berkeley professorwhen did steve leave apple?
• Out of KG mentions
CIKM 2017 Knowledge Graphs: In Theory and Practice 6/47
![Page 19: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/19.jpg)
Entity Linking
Related problems:
• Record linkage/de-duplication in databases• Entity Resolution/name matching• Co-reference resolution, Word Sense disambiguation
CIKM 2017 Knowledge Graphs: In Theory and Practice 7/47
![Page 20: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/20.jpg)
Entity Linking Process
EntityRecognition
Target ListGeneration
Ranking
Named EntityRecognitionWell studied inNLP [17]open sourcesoftware likeStanford NLPtoolkit [16]
Use of dictionaries
Ranking targetentities based on:
• graph basedfeatures
• text/documentbased features
CIKM 2017 Knowledge Graphs: In Theory and Practice 8/47
![Page 21: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/21.jpg)
Entity Linking Process
EntityRecognition
Target ListGeneration
Ranking
Named EntityRecognitionWell studied inNLP [17]open sourcesoftware likeStanford NLPtoolkit [16]
Use of dictionaries
Ranking targetentities based on:
• graph basedfeatures
• text/documentbased features
CIKM 2017 Knowledge Graphs: In Theory and Practice 8/47
![Page 22: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/22.jpg)
Entity Linking Process
EntityRecognition
Target ListGeneration
Ranking
Named EntityRecognitionWell studied inNLP [17]open sourcesoftware likeStanford NLPtoolkit [16]
Use of dictionaries
Ranking targetentities based on:
• graph basedfeatures
• text/documentbased features
CIKM 2017 Knowledge Graphs: In Theory and Practice 8/47
![Page 23: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/23.jpg)
Entity Linking Process
EntityRecognition
Target ListGeneration
Ranking
Named EntityRecognitionWell studied inNLP [17]open sourcesoftware likeStanford NLPtoolkit [16]
Use of dictionaries
Ranking targetentities based on:
• graph basedfeatures
• text/documentbased features
CIKM 2017 Knowledge Graphs: In Theory and Practice 8/47
![Page 24: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/24.jpg)
Entity Linking Process
EntityRecognition
Target ListGeneration
Ranking
Named EntityRecognitionWell studied inNLP [17]open sourcesoftware likeStanford NLPtoolkit [16]
Use of dictionaries
Ranking targetentities based on:
• graph basedfeatures
• text/documentbased features
CIKM 2017 Knowledge Graphs: In Theory and Practice 8/47
![Page 25: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/25.jpg)
Candidate Entity List Generation
• Much of the variation between different entity linkingalgorithms could be explained by quality of candidatesearch components [12]
• Acronym expansions and coreference resolutions lead tosignificant performance gains [12]
• The candidate set should be exhaustive enough but nottoo big to affect efficiency
CIKM 2017 Knowledge Graphs: In Theory and Practice 9/47
![Page 26: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/26.jpg)
Candidate Entity List Generation
• Much of the variation between different entity linkingalgorithms could be explained by quality of candidatesearch components [12]
• Acronym expansions and coreference resolutions lead tosignificant performance gains [12]
• The candidate set should be exhaustive enough but nottoo big to affect efficiency
CIKM 2017 Knowledge Graphs: In Theory and Practice 9/47
![Page 27: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/27.jpg)
Candidate Entity List Generation
• Much of the variation between different entity linkingalgorithms could be explained by quality of candidatesearch components [12]
• Acronym expansions and coreference resolutions lead tosignificant performance gains [12]
• The candidate set should be exhaustive enough but nottoo big to affect efficiency
CIKM 2017 Knowledge Graphs: In Theory and Practice 9/47
![Page 28: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/28.jpg)
Candidate Entity List Generation
• Much of the variation between different entity linkingalgorithms could be explained by quality of candidatesearch components [12]
• Acronym expansions and coreference resolutions lead tosignificant performance gains [12]
• The candidate set should be exhaustive enough but nottoo big to affect efficiency
CIKM 2017 Knowledge Graphs: In Theory and Practice 9/47
![Page 29: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/29.jpg)
Candidate Entity List Generation
Dictionary based MethodsAn offline dictionary of entity names created out of externalsources mapping different possible surface forms of entitynames to their corresponding entities in the KG
• Domain specific sources like Gene name dictionary [18]• Wikipedia/DBPedia
• Page Titles• Disambiguation/Redirect pages• Anchor text of Wikipedia in links
• Anchor text from Web pages to Wikipedia articles• Acronym expansions
CIKM 2017 Knowledge Graphs: In Theory and Practice 10/47
![Page 30: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/30.jpg)
Candidate Entity List Generation
Dictionary based MethodsAn offline dictionary of entity names created out of externalsources mapping different possible surface forms of entitynames to their corresponding entities in the KG
• Domain specific sources like Gene name dictionary [18]
• Wikipedia/DBPedia• Page Titles• Disambiguation/Redirect pages• Anchor text of Wikipedia in links
• Anchor text from Web pages to Wikipedia articles• Acronym expansions
CIKM 2017 Knowledge Graphs: In Theory and Practice 10/47
![Page 31: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/31.jpg)
Candidate Entity List Generation
Dictionary based MethodsAn offline dictionary of entity names created out of externalsources mapping different possible surface forms of entitynames to their corresponding entities in the KG
• Domain specific sources like Gene name dictionary [18]• Wikipedia/DBPedia
• Page Titles• Disambiguation/Redirect pages• Anchor text of Wikipedia in links
• Anchor text from Web pages to Wikipedia articles• Acronym expansions
CIKM 2017 Knowledge Graphs: In Theory and Practice 10/47
![Page 32: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/32.jpg)
Candidate Entity List Generation
Dictionary based MethodsAn offline dictionary of entity names created out of externalsources mapping different possible surface forms of entitynames to their corresponding entities in the KG
• Domain specific sources like Gene name dictionary [18]• Wikipedia/DBPedia
• Page Titles• Disambiguation/Redirect pages• Anchor text of Wikipedia in links
• Anchor text from Web pages to Wikipedia articles
• Acronym expansions
CIKM 2017 Knowledge Graphs: In Theory and Practice 10/47
![Page 33: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/33.jpg)
Candidate Entity List Generation
Dictionary based MethodsAn offline dictionary of entity names created out of externalsources mapping different possible surface forms of entitynames to their corresponding entities in the KG
• Domain specific sources like Gene name dictionary [18]• Wikipedia/DBPedia
• Page Titles• Disambiguation/Redirect pages• Anchor text of Wikipedia in links
• Anchor text from Web pages to Wikipedia articles• Acronym expansions
CIKM 2017 Knowledge Graphs: In Theory and Practice 10/47
![Page 34: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/34.jpg)
Candidate Entity List Generation
Surface Form Entity Canonical Form
Barack Obama < Barack Obama,Person>Barack H. Obama <Barack Obama,Person>USA <United States of America, Country>America <United States of America,Country>Big Apple <New York, City>NYC <New York, City>NY <New York, City>
NY <New York, State>
Simple term match – partial or exact...Obama visited Singapore in 2016...Matches: Barack Obama, Mount Obama, Michelle Obama,..., etc.
CIKM 2017 Knowledge Graphs: In Theory and Practice 11/47
![Page 35: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/35.jpg)
Candidate Entity List Generation
Surface Form Entity Canonical Form
Barack Obama < Barack Obama,Person>Barack H. Obama <Barack Obama,Person>USA <United States of America, Country>America <United States of America,Country>Big Apple <New York, City>NYC <New York, City>NY <New York, City>NY <New York, State>
Simple term match – partial or exact...Obama visited Singapore in 2016...Matches: Barack Obama, Mount Obama, Michelle Obama,..., etc.
CIKM 2017 Knowledge Graphs: In Theory and Practice 11/47
![Page 36: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/36.jpg)
Candidate Entity List Generation
Surface Form Entity Canonical Form
Barack Obama < Barack Obama,Person>Barack H. Obama <Barack Obama,Person>USA <United States of America, Country>America <United States of America,Country>Big Apple <New York, City>NYC <New York, City>NY <New York, City>NY <New York, State>
Simple term match – partial or exact...Obama visited Singapore in 2016...Matches: Barack Obama, Mount Obama, Michelle Obama,..., etc.
CIKM 2017 Knowledge Graphs: In Theory and Practice 11/47
![Page 37: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/37.jpg)
Candidate Entity Ranking
The candidate entity set can be big!For KORE50 dataset:
• 631 candidates on an average per mention in YAGO [23]• 2000+ in Watson KG [4]
Approaches for ranking can be clubbed under two broadcategories:
• Text based• Graph structure based
CIKM 2017 Knowledge Graphs: In Theory and Practice 12/47
![Page 38: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/38.jpg)
Candidate Entity Ranking
The candidate entity set can be big!
For KORE50 dataset:
• 631 candidates on an average per mention in YAGO [23]• 2000+ in Watson KG [4]
Approaches for ranking can be clubbed under two broadcategories:
• Text based• Graph structure based
CIKM 2017 Knowledge Graphs: In Theory and Practice 12/47
![Page 39: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/39.jpg)
Candidate Entity Ranking
The candidate entity set can be big!For KORE50 dataset:
• 631 candidates on an average per mention in YAGO [23]• 2000+ in Watson KG [4]
Approaches for ranking can be clubbed under two broadcategories:
• Text based• Graph structure based
CIKM 2017 Knowledge Graphs: In Theory and Practice 12/47
![Page 40: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/40.jpg)
Candidate Entity Ranking
The candidate entity set can be big!For KORE50 dataset:
• 631 candidates on an average per mention in YAGO [23]• 2000+ in Watson KG [4]
Approaches for ranking can be clubbed under two broadcategories:
• Text based• Graph structure based
CIKM 2017 Knowledge Graphs: In Theory and Practice 12/47
![Page 41: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/41.jpg)
Candidate Entity Ranking
…Obama is in Hawaii this week…{Barack Obama, Michelle Obama, Mt. Obama}
• Similarity between entity name and mention• Term overlap, edit distance, etc.
• Entity Popularity – Wikipedia page views [11, 10]• Wikipedia/web anchor text/ inlinks [20, 13]
…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}Context Matters!
CIKM 2017 Knowledge Graphs: In Theory and Practice 13/47
![Page 42: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/42.jpg)
Candidate Entity Ranking
…Obama is in Hawaii this week…{Barack Obama, Michelle Obama, Mt. Obama}
• Similarity between entity name and mention• Term overlap, edit distance, etc.
• Entity Popularity – Wikipedia page views [11, 10]• Wikipedia/web anchor text/ inlinks [20, 13]
…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}Context Matters!
CIKM 2017 Knowledge Graphs: In Theory and Practice 13/47
![Page 43: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/43.jpg)
Candidate Entity Ranking
…Obama is in Hawaii this week…{Barack Obama, Michelle Obama, Mt. Obama}
• Similarity between entity name and mention• Term overlap, edit distance, etc.
• Entity Popularity – Wikipedia page views [11, 10]
• Wikipedia/web anchor text/ inlinks [20, 13]
…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}Context Matters!
CIKM 2017 Knowledge Graphs: In Theory and Practice 13/47
![Page 44: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/44.jpg)
Candidate Entity Ranking
…Obama is in Hawaii this week…{Barack Obama, Michelle Obama, Mt. Obama}
• Similarity between entity name and mention• Term overlap, edit distance, etc.
• Entity Popularity – Wikipedia page views [11, 10]• Wikipedia/web anchor text/ inlinks [20, 13]
…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}Context Matters!
CIKM 2017 Knowledge Graphs: In Theory and Practice 13/47
![Page 45: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/45.jpg)
Candidate Entity Ranking
…Obama is in Hawaii this week…{Barack Obama, Michelle Obama, Mt. Obama}
• Similarity between entity name and mention• Term overlap, edit distance, etc.
• Entity Popularity – Wikipedia page views [11, 10]• Wikipedia/web anchor text/ inlinks [20, 13]
…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}
Context Matters!
CIKM 2017 Knowledge Graphs: In Theory and Practice 13/47
![Page 46: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/46.jpg)
Candidate Entity Ranking
…Obama is in Hawaii this week…{Barack Obama, Michelle Obama, Mt. Obama}
• Similarity between entity name and mention• Term overlap, edit distance, etc.
• Entity Popularity – Wikipedia page views [11, 10]• Wikipedia/web anchor text/ inlinks [20, 13]
…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}Context Matters!
CIKM 2017 Knowledge Graphs: In Theory and Practice 13/47
![Page 47: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/47.jpg)
Candidate Entity Ranking
Role of Context…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}
• Mention context• text of the document/paragraph in which the mentionappears
• a window of terms around the mention• Entity context representations
• Wikipedia article• Text around anchors• Domain specific models: abstracts of papers containinggene name in titles
Compute similarity between mention and entity contextrepresentations
CIKM 2017 Knowledge Graphs: In Theory and Practice 14/47
![Page 48: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/48.jpg)
Candidate Entity Ranking
Role of Context…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}
• Mention context• text of the document/paragraph in which the mentionappears
• a window of terms around the mention
• Entity context representations• Wikipedia article• Text around anchors• Domain specific models: abstracts of papers containinggene name in titles
Compute similarity between mention and entity contextrepresentations
CIKM 2017 Knowledge Graphs: In Theory and Practice 14/47
![Page 49: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/49.jpg)
Candidate Entity Ranking
Role of Context…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}
• Mention context• text of the document/paragraph in which the mentionappears
• a window of terms around the mention• Entity context representations
• Wikipedia article• Text around anchors• Domain specific models: abstracts of papers containinggene name in titles
Compute similarity between mention and entity contextrepresentations
CIKM 2017 Knowledge Graphs: In Theory and Practice 14/47
![Page 50: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/50.jpg)
Candidate Entity Ranking
Role of Context…when did Steve leave apple…{Steve Jobs,Steve Wozniak,Steve Ballmer}
• Mention context• text of the document/paragraph in which the mentionappears
• a window of terms around the mention• Entity context representations
• Wikipedia article• Text around anchors• Domain specific models: abstracts of papers containinggene name in titles
Compute similarity between mention and entity contextrepresentations
CIKM 2017 Knowledge Graphs: In Theory and Practice 14/47
![Page 51: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/51.jpg)
Candidate Entity Ranking
Graph Based Features Focus on strength between entities,often useful in collective entity linking
• Simplest graph based measure – Entity Popularity
pop(e) = nbrCount(e)∑e′∈E
nbrCount(e′)(1)
In Wikipedia graph, inlinks and outlinks can be used tocompute popularity
Next we review some measures useful for collective entitylinking
CIKM 2017 Knowledge Graphs: In Theory and Practice 15/47
![Page 52: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/52.jpg)
Candidate Entity Ranking
Graph Based Features Focus on strength between entities,often useful in collective entity linking
• Simplest graph based measure – Entity Popularity
pop(e) = nbrCount(e)∑e′∈E
nbrCount(e′)(1)
In Wikipedia graph, inlinks and outlinks can be used tocompute popularity
Next we review some measures useful for collective entitylinking
CIKM 2017 Knowledge Graphs: In Theory and Practice 15/47
![Page 53: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/53.jpg)
Candidate Entity Ranking
Graph Based Features Focus on strength between entities,often useful in collective entity linking
• Simplest graph based measure – Entity Popularity
pop(e) = nbrCount(e)∑e′∈E
nbrCount(e′)(1)
In Wikipedia graph, inlinks and outlinks can be used tocompute popularity
Next we review some measures useful for collective entitylinking
CIKM 2017 Knowledge Graphs: In Theory and Practice 15/47
![Page 54: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/54.jpg)
Candidate Entity Ranking
Linking/Resolving/Disambiguating Multiple Entitiessimultaneously
Image Source: [26]CIKM 2017 Knowledge Graphs: In Theory and Practice 16/47
![Page 55: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/55.jpg)
Candidate Entity Ranking
Brad and Angelina were holidaying in Paris.
• Jaccard IndexJ(a,b) = |A ∩ B|
|A ∪ B|(2)
• Milne-Witten Similarity [26]
MW(a,b) = log(max(|A|, |B|))− log(|A ∩ B|)log(|N |)− log(min(|A|, |B|))
(3)
where, A and B are the set of neighbors of entities a and b,respectively.
• Adamic Adar [1]
AA(a,b) =∑n∈A∪B
log( 1degree(n)
) (4)
CIKM 2017 Knowledge Graphs: In Theory and Practice 17/47
![Page 56: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/56.jpg)
Candidate Entity Ranking
Brad and Angelina were holidaying in Paris.
• Jaccard IndexJ(a,b) = |A ∩ B|
|A ∪ B|(2)
• Milne-Witten Similarity [26]
MW(a,b) = log(max(|A|, |B|))− log(|A ∩ B|)log(|N |)− log(min(|A|, |B|))
(3)
where, A and B are the set of neighbors of entities a and b,respectively.
• Adamic Adar [1]
AA(a,b) =∑n∈A∪B
log( 1degree(n)
) (4)
CIKM 2017 Knowledge Graphs: In Theory and Practice 17/47
![Page 57: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/57.jpg)
Candidate Entity Ranking
Brad and Angelina were holidaying in Paris.
• Jaccard IndexJ(a,b) = |A ∩ B|
|A ∪ B|(2)
• Milne-Witten Similarity [26]
MW(a,b) = log(max(|A|, |B|))− log(|A ∩ B|)log(|N |)− log(min(|A|, |B|))
(3)
where, A and B are the set of neighbors of entities a and b,respectively.
• Adamic Adar [1]
AA(a,b) =∑n∈A∪B
log( 1degree(n)
) (4)
CIKM 2017 Knowledge Graphs: In Theory and Practice 17/47
![Page 58: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/58.jpg)
Candidate Entity Ranking
Brad and Angelina were holidaying in Paris.
• Jaccard IndexJ(a,b) = |A ∩ B|
|A ∪ B|(2)
• Milne-Witten Similarity [26]
MW(a,b) = log(max(|A|, |B|))− log(|A ∩ B|)log(|N |)− log(min(|A|, |B|))
(3)
where, A and B are the set of neighbors of entities a and b,respectively.
• Adamic Adar [1]
AA(a,b) =∑n∈A∪B
log( 1degree(n)
) (4)
CIKM 2017 Knowledge Graphs: In Theory and Practice 17/47
![Page 59: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/59.jpg)
Candidate Entity Ranking
• These features can be used in supervised or unsupervisedsettings
• Choice of features depend on data/domain at hand. Manyfeatures are specific for Wikipedia, that may not beapplicable to other textual data.
• Trade off between accuracy and efficiency while designingyour systems
CIKM 2017 Knowledge Graphs: In Theory and Practice 18/47
![Page 60: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/60.jpg)
Entity Linking as implemented in Watson KG1
Which search algorithm did Sergey and Larry invent?
1S. Bhatia and A. Jain. “Context Sensitive Entity Linking of Search Queries in Enterprise Knowledge Graphs”. In:International Semantic Web Conference. Springer. 2016, pp. 50–54.
CIKM 2017 Knowledge Graphs: In Theory and Practice 19/47
![Page 61: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/61.jpg)
Entity Linking as implemented in Watson KG1
Which search algorithm did Sergey and Larry invent?
1S. Bhatia and A. Jain. “Context Sensitive Entity Linking of Search Queries in Enterprise Knowledge Graphs”. In:International Semantic Web Conference. Springer. 2016, pp. 50–54.
CIKM 2017 Knowledge Graphs: In Theory and Practice 19/47
![Page 62: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/62.jpg)
Entity Exploration
![Page 63: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/63.jpg)
Entity Exploration
We found the entity of interest.
Knowing more about the entity
• Finding entities related to entity of interest• Properties of entities• Going beyond immediate neighborhood of the entity
CIKM 2017 Knowledge Graphs: In Theory and Practice 20/47
![Page 64: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/64.jpg)
Entity Retrieval
• Entity Box in web queries
• Lots of useful informationabout the query entity
• ≈ 40% of all web queries areentity queries [19]
• Many QA queries can beanswered by the underlyingKnowledge Base
CIKM 2017 Knowledge Graphs: In Theory and Practice 21/47
![Page 65: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/65.jpg)
Entity Retrieval
Related Entity Finding track at TREC [3]
Input: Entity Name and Search IntentOutput: Ranked list of entity documents – entities embeddedin documentsExample:
Query: BlackberryIntent:Carriers that carry Blackberry phonesExample Answers:Verizon, AT&T, etc.
CIKM 2017 Knowledge Graphs: In Theory and Practice 22/47
![Page 66: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/66.jpg)
Entity Retrieval
Related Entity Finding track at TREC [3]Input: Entity Name and Search Intent
Output: Ranked list of entity documents – entities embeddedin documentsExample:
Query: BlackberryIntent:Carriers that carry Blackberry phonesExample Answers:Verizon, AT&T, etc.
CIKM 2017 Knowledge Graphs: In Theory and Practice 22/47
![Page 67: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/67.jpg)
Entity Retrieval
Related Entity Finding track at TREC [3]Input: Entity Name and Search IntentOutput: Ranked list of entity documents – entities embeddedin documents
Example:
Query: BlackberryIntent:Carriers that carry Blackberry phonesExample Answers:Verizon, AT&T, etc.
CIKM 2017 Knowledge Graphs: In Theory and Practice 22/47
![Page 68: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/68.jpg)
Entity Retrieval
Related Entity Finding track at TREC [3]Input: Entity Name and Search IntentOutput: Ranked list of entity documents – entities embeddedin documentsExample:
Query: BlackberryIntent:Carriers that carry Blackberry phonesExample Answers:Verizon, AT&T, etc.
CIKM 2017 Knowledge Graphs: In Theory and Practice 22/47
![Page 69: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/69.jpg)
Entity Retrieval
Components of Related Entity Ranking [7]2
For a given input entity es, type T of target entity, and a relationdescription R, we wish to rank the target entities as follows:
P(e|es, T,R) ∝ P(R|es, e)︸ ︷︷ ︸Context Modeling
× P(e|es)︸ ︷︷ ︸Co-occurrence
× P(T|e)︸ ︷︷ ︸Type Filtering
(5)
query Co-occurrence Type Filter Context Modeling results
2M. Bron, K. Balog, and M. De Rijke. “Ranking related entities: components and analyses”. In: Proceedings of the19th ACM international conference on Information and knowledge management. ACM. 2010, pp. 1079–1088.
CIKM 2017 Knowledge Graphs: In Theory and Practice 23/47
![Page 70: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/70.jpg)
Entity Retrieval
Components of Related Entity Ranking [7]2
For a given input entity es, type T of target entity, and a relationdescription R, we wish to rank the target entities as follows:
P(e|es, T,R) ∝ P(R|es, e)︸ ︷︷ ︸Context Modeling
× P(e|es)︸ ︷︷ ︸Co-occurrence
× P(T|e)︸ ︷︷ ︸Type Filtering
(5)
query Co-occurrence Type Filter Context Modeling results
2M. Bron, K. Balog, and M. De Rijke. “Ranking related entities: components and analyses”. In: Proceedings of the19th ACM international conference on Information and knowledge management. ACM. 2010, pp. 1079–1088.
CIKM 2017 Knowledge Graphs: In Theory and Practice 23/47
![Page 71: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/71.jpg)
Entity Retrieval
Components of Related Entity Ranking [7]2
For a given input entity es, type T of target entity, and a relationdescription R, we wish to rank the target entities as follows:
P(e|es, T,R) ∝ P(R|es, e)︸ ︷︷ ︸Context Modeling
× P(e|es)︸ ︷︷ ︸Co-occurrence
× P(T|e)︸ ︷︷ ︸Type Filtering
(5)
query Co-occurrence Type Filter Context Modeling results
2M. Bron, K. Balog, and M. De Rijke. “Ranking related entities: components and analyses”. In: Proceedings of the19th ACM international conference on Information and knowledge management. ACM. 2010, pp. 1079–1088.
CIKM 2017 Knowledge Graphs: In Theory and Practice 23/47
![Page 72: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/72.jpg)
Entity Retrieval
Co-occurrenceP(e|es) =
cooc(e, es)∑e′∈E
cooc(e′, es)
Type Filtering
• Wikipedia categories• Named entity recognizer tools
Context ModelingCo-occurrence language model Θees approximated bydocuments in which e, Es co-occur
P(R|e, es) =∏t∈R
P(t|Θees)
CIKM 2017 Knowledge Graphs: In Theory and Practice 24/47
![Page 73: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/73.jpg)
Entity Retrieval
Co-occurrenceP(e|es) =
cooc(e, es)∑e′∈E
cooc(e′, es)
Type Filtering
• Wikipedia categories• Named entity recognizer tools
Context ModelingCo-occurrence language model Θees approximated bydocuments in which e, Es co-occur
P(R|e, es) =∏t∈R
P(t|Θees)
CIKM 2017 Knowledge Graphs: In Theory and Practice 24/47
![Page 74: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/74.jpg)
Entity Retrieval
Co-occurrenceP(e|es) =
cooc(e, es)∑e′∈E
cooc(e′, es)
Type Filtering
• Wikipedia categories• Named entity recognizer tools
Context ModelingCo-occurrence language model Θees approximated bydocuments in which e, Es co-occur
P(R|e, es) =∏t∈R
P(t|Θees)
CIKM 2017 Knowledge Graphs: In Theory and Practice 24/47
![Page 75: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/75.jpg)
Entity Recommendation for Web Queries I
Entity recommendations for web search queries[6]3
• Co-occurrence features• query logs, user sessions• flickr and twitter tags
• frequency• Graph theoretic features
• Page rank on entity graph• Common neighbors between two entities
3R. Blanco et al. “Entity recommendations in web search”. In: International Semantic Web Conference. Springer.2013, pp. 33–48.
CIKM 2017 Knowledge Graphs: In Theory and Practice 25/47
![Page 76: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/76.jpg)
Entity Recommendation for Web Queries I
Entity recommendations for web search queries[6]3
• Co-occurrence features• query logs, user sessions• flickr and twitter tags
• frequency• Graph theoretic features
• Page rank on entity graph• Common neighbors between two entities
3R. Blanco et al. “Entity recommendations in web search”. In: International Semantic Web Conference. Springer.2013, pp. 33–48.
CIKM 2017 Knowledge Graphs: In Theory and Practice 25/47
![Page 77: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/77.jpg)
Entity Recommendation for Web Queries I
Entity recommendations for web search queries[6]3
• Co-occurrence features• query logs, user sessions• flickr and twitter tags
• frequency
• Graph theoretic features• Page rank on entity graph• Common neighbors between two entities
3R. Blanco et al. “Entity recommendations in web search”. In: International Semantic Web Conference. Springer.2013, pp. 33–48.
CIKM 2017 Knowledge Graphs: In Theory and Practice 25/47
![Page 78: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/78.jpg)
Entity Recommendation for Web Queries I
Entity recommendations for web search queries[6]3
• Co-occurrence features• query logs, user sessions• flickr and twitter tags
• frequency• Graph theoretic features
• Page rank on entity graph• Common neighbors between two entities
3R. Blanco et al. “Entity recommendations in web search”. In: International Semantic Web Conference. Springer.2013, pp. 33–48.
CIKM 2017 Knowledge Graphs: In Theory and Practice 25/47
![Page 79: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/79.jpg)
Entity Recommendation for Web Queries II
Learning to rank using text and graph based features[21]4
• Given a web query, retrieve relevant documents,• Identify entities present in them using entity linkingmethods
• Rank these entities using graph theoretic and text basedfeatures
• Reformulates entity retrieval/recommendation as ad hocdocument retrieval
4M. Schuhmacher, L. Dietz, and S. Paolo Ponzetto. “Ranking entities for web queries through text and knowledge”.In: Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. ACM.2015, pp. 1461–1470.
CIKM 2017 Knowledge Graphs: In Theory and Practice 26/47
![Page 80: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/80.jpg)
Entity Recommendation for Web Queries II
Learning to rank using text and graph based features[21]4
• Given a web query, retrieve relevant documents,
• Identify entities present in them using entity linkingmethods
• Rank these entities using graph theoretic and text basedfeatures
• Reformulates entity retrieval/recommendation as ad hocdocument retrieval
4M. Schuhmacher, L. Dietz, and S. Paolo Ponzetto. “Ranking entities for web queries through text and knowledge”.In: Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. ACM.2015, pp. 1461–1470.
CIKM 2017 Knowledge Graphs: In Theory and Practice 26/47
![Page 81: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/81.jpg)
Entity Recommendation for Web Queries II
Learning to rank using text and graph based features[21]4
• Given a web query, retrieve relevant documents,• Identify entities present in them using entity linkingmethods
• Rank these entities using graph theoretic and text basedfeatures
• Reformulates entity retrieval/recommendation as ad hocdocument retrieval
4M. Schuhmacher, L. Dietz, and S. Paolo Ponzetto. “Ranking entities for web queries through text and knowledge”.In: Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. ACM.2015, pp. 1461–1470.
CIKM 2017 Knowledge Graphs: In Theory and Practice 26/47
![Page 82: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/82.jpg)
Entity Recommendation for Web Queries II
Learning to rank using text and graph based features[21]4
• Given a web query, retrieve relevant documents,• Identify entities present in them using entity linkingmethods
• Rank these entities using graph theoretic and text basedfeatures
• Reformulates entity retrieval/recommendation as ad hocdocument retrieval
4M. Schuhmacher, L. Dietz, and S. Paolo Ponzetto. “Ranking entities for web queries through text and knowledge”.In: Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. ACM.2015, pp. 1461–1470.
CIKM 2017 Knowledge Graphs: In Theory and Practice 26/47
![Page 83: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/83.jpg)
Entity Recommendation for Web Queries II
Learning to rank using text and graph based features[21]4
• Given a web query, retrieve relevant documents,• Identify entities present in them using entity linkingmethods
• Rank these entities using graph theoretic and text basedfeatures
• Reformulates entity retrieval/recommendation as ad hocdocument retrieval
4M. Schuhmacher, L. Dietz, and S. Paolo Ponzetto. “Ranking entities for web queries through text and knowledge”.In: Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. ACM.2015, pp. 1461–1470.
CIKM 2017 Knowledge Graphs: In Theory and Practice 26/47
![Page 84: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/84.jpg)
Entity Exploration
Till now, we have focused on finding entities
Let us focus our attention now on finding about entities
UnitedStates
FloridaFrance
BarackObama
Washington
AbrahamLincoln
Hollywood
SiliconValley
Relationships of similar types can be clustered and thenexplored based on user requirements [27]
CIKM 2017 Knowledge Graphs: In Theory and Practice 27/47
![Page 85: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/85.jpg)
Entity Exploration
Till now, we have focused on finding entitiesLet us focus our attention now on finding about entities
UnitedStates
FloridaFrance
BarackObama
Washington
AbrahamLincoln
Hollywood
SiliconValley
Relationships of similar types can be clustered and thenexplored based on user requirements [27]
CIKM 2017 Knowledge Graphs: In Theory and Practice 27/47
![Page 86: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/86.jpg)
Entity Exploration
Till now, we have focused on finding entitiesLet us focus our attention now on finding about entities
UnitedStates
FloridaFrance
BarackObama
Washington
AbrahamLincoln
Hollywood
SiliconValley
Relationships of similar types can be clustered and thenexplored based on user requirements [27]
CIKM 2017 Knowledge Graphs: In Theory and Practice 27/47
![Page 87: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/87.jpg)
Entity Exploration
Till now, we have focused on finding entitiesLet us focus our attention now on finding about entities
UnitedStates
FloridaFrance
BarackObama
Washington
AbrahamLincoln
Hollywood
SiliconValley
Relationships of similar types can be clustered and thenexplored based on user requirements [27]
CIKM 2017 Knowledge Graphs: In Theory and Practice 27/47
![Page 88: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/88.jpg)
Entity Exploration – Fact Ranking
What are the most important facts about an entity?5
Given asource entity es, we wish to compute the probability P(r, et|es)
P(r, et|es) ∝ P(et)︸ ︷︷ ︸entity prior
× P(es|et)︸ ︷︷ ︸entity affinity
× P(r|es, et)︸ ︷︷ ︸relationship strength
(6)
Entity Prior:P(et) ∝ relCount(et) (7)
Entity Affinity
P(e|et) =∑
ri∈R(es,et) w(ri)× ri∑ri∈R(et) w(ri)× ri
(8)
Relationship Strength
P(r|es, et) =mentionCount(r, es, et)∑
r∈R(es,et)mentionCount(r, es, et)(9)
5S. Bhatia et al. “Separating Wheat from the Chaff–A Relationship Ranking Algorithm”. In: International SemanticWeb Conference. Springer. 2016, pp. 79–83.
CIKM 2017 Knowledge Graphs: In Theory and Practice 28/47
![Page 89: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/89.jpg)
Entity Exploration – Fact Ranking
What are the most important facts about an entity?5 Given asource entity es, we wish to compute the probability P(r, et|es)
P(r, et|es) ∝ P(et)︸ ︷︷ ︸entity prior
× P(es|et)︸ ︷︷ ︸entity affinity
× P(r|es, et)︸ ︷︷ ︸relationship strength
(6)
Entity Prior:P(et) ∝ relCount(et) (7)
Entity Affinity
P(e|et) =∑
ri∈R(es,et) w(ri)× ri∑ri∈R(et) w(ri)× ri
(8)
Relationship Strength
P(r|es, et) =mentionCount(r, es, et)∑
r∈R(es,et)mentionCount(r, es, et)(9)
5S. Bhatia et al. “Separating Wheat from the Chaff–A Relationship Ranking Algorithm”. In: International SemanticWeb Conference. Springer. 2016, pp. 79–83.
CIKM 2017 Knowledge Graphs: In Theory and Practice 28/47
![Page 90: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/90.jpg)
Entity Exploration – Fact Ranking
What are the most important facts about an entity?5 Given asource entity es, we wish to compute the probability P(r, et|es)
P(r, et|es) ∝ P(et)︸ ︷︷ ︸entity prior
× P(es|et)︸ ︷︷ ︸entity affinity
× P(r|es, et)︸ ︷︷ ︸relationship strength
(6)
Entity Prior:P(et) ∝ relCount(et) (7)
Entity Affinity
P(e|et) =∑
ri∈R(es,et) w(ri)× ri∑ri∈R(et) w(ri)× ri
(8)
Relationship Strength
P(r|es, et) =mentionCount(r, es, et)∑
r∈R(es,et)mentionCount(r, es, et)(9)
5S. Bhatia et al. “Separating Wheat from the Chaff–A Relationship Ranking Algorithm”. In: International SemanticWeb Conference. Springer. 2016, pp. 79–83.
CIKM 2017 Knowledge Graphs: In Theory and Practice 28/47
![Page 91: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/91.jpg)
Entity Exploration – Fact Ranking
What are the most important facts about an entity?5 Given asource entity es, we wish to compute the probability P(r, et|es)
P(r, et|es) ∝ P(et)︸ ︷︷ ︸entity prior
× P(es|et)︸ ︷︷ ︸entity affinity
× P(r|es, et)︸ ︷︷ ︸relationship strength
(6)
Entity Prior:P(et) ∝ relCount(et) (7)
Entity Affinity
P(e|et) =∑
ri∈R(es,et) w(ri)× ri∑ri∈R(et) w(ri)× ri
(8)
Relationship Strength
P(r|es, et) =mentionCount(r, es, et)∑
r∈R(es,et)mentionCount(r, es, et)(9)
5S. Bhatia et al. “Separating Wheat from the Chaff–A Relationship Ranking Algorithm”. In: International SemanticWeb Conference. Springer. 2016, pp. 79–83.
CIKM 2017 Knowledge Graphs: In Theory and Practice 28/47
![Page 92: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/92.jpg)
Entity Exploration – Fact Ranking
What are the most important facts about an entity?5 Given asource entity es, we wish to compute the probability P(r, et|es)
P(r, et|es) ∝ P(et)︸ ︷︷ ︸entity prior
× P(es|et)︸ ︷︷ ︸entity affinity
× P(r|es, et)︸ ︷︷ ︸relationship strength
(6)
Entity Prior:P(et) ∝ relCount(et) (7)
Entity Affinity
P(e|et) =∑
ri∈R(es,et) w(ri)× ri∑ri∈R(et) w(ri)× ri
(8)
Relationship Strength
P(r|es, et) =mentionCount(r, es, et)∑
r∈R(es,et)mentionCount(r, es, et)(9)
5S. Bhatia et al. “Separating Wheat from the Chaff–A Relationship Ranking Algorithm”. In: International SemanticWeb Conference. Springer. 2016, pp. 79–83.
CIKM 2017 Knowledge Graphs: In Theory and Practice 28/47
![Page 93: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/93.jpg)
Entity Exploration - Fact Ranking
CIKM 2017 Knowledge Graphs: In Theory and Practice 29/47
![Page 94: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/94.jpg)
Entity Exploration – Moving Beyond the Neighborhood
Till now, we have limited our attention to relations of theentity and it’s immediate neighborhood.
What lies after that?
CIKM 2017 Knowledge Graphs: In Theory and Practice 30/47
![Page 95: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/95.jpg)
Entity Exploration – Moving Beyond the Neighborhood
Till now, we have limited our attention to relations of theentity and it’s immediate neighborhood.What lies after that?
CIKM 2017 Knowledge Graphs: In Theory and Practice 30/47
![Page 96: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/96.jpg)
Entity Exploration – Moving Beyond the Neighborhood
Discovering and Explaining Higher Order Relations BetweenEntities
Can we tell how are they connected?
CIKM 2017 Knowledge Graphs: In Theory and Practice 31/47
![Page 97: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/97.jpg)
Entity Exploration – Moving Beyond the Neighborhood
Discovering and Explaining Higher Order Relations BetweenEntities
Can we tell how are they connected?
CIKM 2017 Knowledge Graphs: In Theory and Practice 31/47
![Page 98: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/98.jpg)
Entity Exploration – Moving Beyond the Neighborhood
Discovering and Explaining Higher Order Relations BetweenEntities
Can we tell how are they connected?
CIKM 2017 Knowledge Graphs: In Theory and Practice 31/47
![Page 99: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/99.jpg)
Entity Exploration – Moving Beyond the Neighborhood
Discovering and Explaining Higher Order Relations BetweenEntities
Can we tell how are they connected?
CIKM 2017 Knowledge Graphs: In Theory and Practice 31/47
![Page 100: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/100.jpg)
Path Ranking
• Thousands of such paths• Too generic – obvious relations
CIKM 2017 Knowledge Graphs: In Theory and Practice 32/47
![Page 101: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/101.jpg)
Path Ranking
• Thousands of such paths• Too generic – obvious relations
CIKM 2017 Knowledge Graphs: In Theory and Practice 32/47
![Page 102: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/102.jpg)
Path Ranking
Three components for ranking possible paths [2]
Specificity: Popular entities given lower scores
spec(p) =∑e∈p
spec(e);where: spec(e) = log(1+ 1/docCount(e)) (10)
Reduces generic paths, but boosts noise entities
Connectivity: A strongly connected path consists of strong edges.
score(ea, eb) = ~dea · ~deb (11)
Cohesiveness:
score(p) =n−1∑i=2
score(ei) =n−1∑i=2
~dei−1 · ~dei+1 (12)
CIKM 2017 Knowledge Graphs: In Theory and Practice 33/47
![Page 103: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/103.jpg)
Path Ranking
Three components for ranking possible paths [2]Specificity: Popular entities given lower scores
spec(p) =∑e∈p
spec(e);where: spec(e) = log(1+ 1/docCount(e)) (10)
Reduces generic paths, but boosts noise entities
Connectivity: A strongly connected path consists of strong edges.
score(ea, eb) = ~dea · ~deb (11)
Cohesiveness:
score(p) =n−1∑i=2
score(ei) =n−1∑i=2
~dei−1 · ~dei+1 (12)
CIKM 2017 Knowledge Graphs: In Theory and Practice 33/47
![Page 104: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/104.jpg)
Path Ranking
Three components for ranking possible paths [2]Specificity: Popular entities given lower scores
spec(p) =∑e∈p
spec(e);where: spec(e) = log(1+ 1/docCount(e)) (10)
Reduces generic paths, but boosts noise entities
Connectivity: A strongly connected path consists of strong edges.
score(ea, eb) = ~dea · ~deb (11)
Cohesiveness:
score(p) =n−1∑i=2
score(ei) =n−1∑i=2
~dei−1 · ~dei+1 (12)
CIKM 2017 Knowledge Graphs: In Theory and Practice 33/47
![Page 105: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/105.jpg)
Path Ranking
Three components for ranking possible paths [2]Specificity: Popular entities given lower scores
spec(p) =∑e∈p
spec(e);where: spec(e) = log(1+ 1/docCount(e)) (10)
Reduces generic paths, but boosts noise entities
Connectivity: A strongly connected path consists of strong edges.
score(ea, eb) = ~dea · ~deb (11)
Cohesiveness:
score(p) =n−1∑i=2
score(ei) =n−1∑i=2
~dei−1 · ~dei+1 (12)
CIKM 2017 Knowledge Graphs: In Theory and Practice 33/47
![Page 106: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/106.jpg)
Path Ranking
CIKM 2017 Knowledge Graphs: In Theory and Practice 34/47
![Page 107: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/107.jpg)
Path Ranking
CIKM 2017 Knowledge Graphs: In Theory and Practice 35/47
![Page 108: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/108.jpg)
Application Example from Life Sciences
Predicting Drug-Drug Interactions(DDI)6
• DDI are a major cause of preventable adverse drugreactions
• Clinical studies can not accurately determine all possibleDDIs
• Can we utilize knowledge about drugs to predict possibleDDIs?
6A. Fokoue et al. “Predicting drug-drug interactions through large-scale similarity-based link prediction”. In:International Semantic Web Conference. Springer. 2016, pp. 774–789.
CIKM 2017 Knowledge Graphs: In Theory and Practice 36/47
![Page 109: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/109.jpg)
Application Example from Life Sciences
Predicting Drug-Drug Interactions(DDI)6
• DDI are a major cause of preventable adverse drugreactions
• Clinical studies can not accurately determine all possibleDDIs
• Can we utilize knowledge about drugs to predict possibleDDIs?
6A. Fokoue et al. “Predicting drug-drug interactions through large-scale similarity-based link prediction”. In:International Semantic Web Conference. Springer. 2016, pp. 774–789.
CIKM 2017 Knowledge Graphs: In Theory and Practice 36/47
![Page 110: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/110.jpg)
Application Example from Life Sciences
Predicting Drug-Drug Interactions(DDI)6
• DDI are a major cause of preventable adverse drugreactions
• Clinical studies can not accurately determine all possibleDDIs
• Can we utilize knowledge about drugs to predict possibleDDIs?
6A. Fokoue et al. “Predicting drug-drug interactions through large-scale similarity-based link prediction”. In:International Semantic Web Conference. Springer. 2016, pp. 774–789.
CIKM 2017 Knowledge Graphs: In Theory and Practice 36/47
![Page 111: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/111.jpg)
Application Example from Life Sciences
Predicting Drug-Drug Interactions(DDI)6
• DDI are a major cause of preventable adverse drugreactions
• Clinical studies can not accurately determine all possibleDDIs
• Can we utilize knowledge about drugs to predict possibleDDIs?
6A. Fokoue et al. “Predicting drug-drug interactions through large-scale similarity-based link prediction”. In:International Semantic Web Conference. Springer. 2016, pp. 774–789.
CIKM 2017 Knowledge Graphs: In Theory and Practice 36/47
![Page 112: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/112.jpg)
Application Example from Life Sciences
Create a KG out of existing information about drugs and theirinteractions with genes, enzymes, molecules, etc.
CIKM 2017 Knowledge Graphs: In Theory and Practice 37/47
![Page 113: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/113.jpg)
Application Example from Life Sciences
• Given a pair of drugs, extract features based onphysiological effect, side effect, targets, drug targets,chemical structure, etc.
• Perform supervised classification using logistic regression• Retrospective Analysis: Known DDIs til January 2011 astraining.
• Could predict ≈ 68% of DDIs discovered after January 2011till December 2014.
CIKM 2017 Knowledge Graphs: In Theory and Practice 38/47
![Page 114: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/114.jpg)
Future Research Directions
• Reasoning over Knowledge Graphs
• KG Completion [8, 22, 15]• Complex QA Systems
• Explaining relations present in a graph [24, 14]• Graph and text joint modeling [25, 28]• Ask domain experts!
CIKM 2017 Knowledge Graphs: In Theory and Practice 39/47
![Page 115: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/115.jpg)
Future Research Directions
• Reasoning over Knowledge Graphs
• KG Completion [8, 22, 15]• Complex QA Systems
• Explaining relations present in a graph [24, 14]• Graph and text joint modeling [25, 28]• Ask domain experts!
CIKM 2017 Knowledge Graphs: In Theory and Practice 39/47
![Page 116: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/116.jpg)
Future Research Directions
• Reasoning over Knowledge Graphs
• KG Completion [8, 22, 15]• Complex QA Systems
• Explaining relations present in a graph [24, 14]
• Graph and text joint modeling [25, 28]• Ask domain experts!
CIKM 2017 Knowledge Graphs: In Theory and Practice 39/47
![Page 117: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/117.jpg)
Future Research Directions
• Reasoning over Knowledge Graphs
• KG Completion [8, 22, 15]• Complex QA Systems
• Explaining relations present in a graph [24, 14]• Graph and text joint modeling [25, 28]
• Ask domain experts!
CIKM 2017 Knowledge Graphs: In Theory and Practice 39/47
![Page 118: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/118.jpg)
Future Research Directions
• Reasoning over Knowledge Graphs
• KG Completion [8, 22, 15]• Complex QA Systems
• Explaining relations present in a graph [24, 14]• Graph and text joint modeling [25, 28]• Ask domain experts!
CIKM 2017 Knowledge Graphs: In Theory and Practice 39/47
![Page 119: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/119.jpg)
DEMO
CIKM 2017 Knowledge Graphs: In Theory and Practice 40/47
![Page 120: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/120.jpg)
Conclusions
• KG can provide structure to your unstructured data!
• We wanted to provide an overview of tools/techniquesthat have worked well in the past, and challenges you mayface
• Should help you get started with a pretty strong baselinesystem
• Be careful in selecting the KG appropriate for your domainand requirements.
• Keep in mind the scale and efficiency issues• You will have to work with lots of noisy and erroneous data• But the efforts required are worth it!
CIKM 2017 Knowledge Graphs: In Theory and Practice 41/47
![Page 121: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/121.jpg)
Conclusions
• KG can provide structure to your unstructured data!• We wanted to provide an overview of tools/techniquesthat have worked well in the past, and challenges you mayface
• Should help you get started with a pretty strong baselinesystem
• Be careful in selecting the KG appropriate for your domainand requirements.
• Keep in mind the scale and efficiency issues• You will have to work with lots of noisy and erroneous data• But the efforts required are worth it!
CIKM 2017 Knowledge Graphs: In Theory and Practice 41/47
![Page 122: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/122.jpg)
Conclusions
• KG can provide structure to your unstructured data!• We wanted to provide an overview of tools/techniquesthat have worked well in the past, and challenges you mayface
• Should help you get started with a pretty strong baselinesystem
• Be careful in selecting the KG appropriate for your domainand requirements.
• Keep in mind the scale and efficiency issues• You will have to work with lots of noisy and erroneous data• But the efforts required are worth it!
CIKM 2017 Knowledge Graphs: In Theory and Practice 41/47
![Page 123: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/123.jpg)
Conclusions
• KG can provide structure to your unstructured data!• We wanted to provide an overview of tools/techniquesthat have worked well in the past, and challenges you mayface
• Should help you get started with a pretty strong baselinesystem
• Be careful in selecting the KG appropriate for your domainand requirements.
• Keep in mind the scale and efficiency issues• You will have to work with lots of noisy and erroneous data• But the efforts required are worth it!
CIKM 2017 Knowledge Graphs: In Theory and Practice 41/47
![Page 124: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/124.jpg)
Conclusions
• KG can provide structure to your unstructured data!• We wanted to provide an overview of tools/techniquesthat have worked well in the past, and challenges you mayface
• Should help you get started with a pretty strong baselinesystem
• Be careful in selecting the KG appropriate for your domainand requirements.
• Keep in mind the scale and efficiency issues
• You will have to work with lots of noisy and erroneous data• But the efforts required are worth it!
CIKM 2017 Knowledge Graphs: In Theory and Practice 41/47
![Page 125: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/125.jpg)
Conclusions
• KG can provide structure to your unstructured data!• We wanted to provide an overview of tools/techniquesthat have worked well in the past, and challenges you mayface
• Should help you get started with a pretty strong baselinesystem
• Be careful in selecting the KG appropriate for your domainand requirements.
• Keep in mind the scale and efficiency issues• You will have to work with lots of noisy and erroneous data
• But the efforts required are worth it!
CIKM 2017 Knowledge Graphs: In Theory and Practice 41/47
![Page 126: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/126.jpg)
Conclusions
• KG can provide structure to your unstructured data!• We wanted to provide an overview of tools/techniquesthat have worked well in the past, and challenges you mayface
• Should help you get started with a pretty strong baselinesystem
• Be careful in selecting the KG appropriate for your domainand requirements.
• Keep in mind the scale and efficiency issues• You will have to work with lots of noisy and erroneous data• But the efforts required are worth it!
CIKM 2017 Knowledge Graphs: In Theory and Practice 41/47
![Page 127: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/127.jpg)
Thanks!!!Suggestions and Questions Welcome!
Slides available at http://sumitbhatia.net/source/knowledge-graph-tutorial.html
CIKM 2017 Knowledge Graphs: In Theory and Practice 42/47
![Page 128: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/128.jpg)
References i
[1] L. A. Adamic and E. Adar. “Friends and neighbors on the web”. In: Socialnetworks 25.3 (2003), pp. 211–230.
[2] N. Aggarwal, S. Bhatia, and V. Misra. “Connecting the Dots: ExplainingRelationships Between Unconnected Entities in a Knowledge Graph”. In:International Semantic Web Conference. Springer. 2016, pp. 35–39.
[3] K. Balog et al. “Overview of the TREC 2009 entity track”. In: In Proceedings of theEighteenth Text REtrieval Conference. 2009.
[4] S. Bhatia and A. Jain. “Context Sensitive Entity Linking of Search Queries inEnterprise Knowledge Graphs”. In: International Semantic Web Conference.Springer. 2016, pp. 50–54.
[5] S. Bhatia et al. “Separating Wheat from the Chaff–A Relationship RankingAlgorithm”. In: International Semantic Web Conference. Springer. 2016, pp. 79–83.
[6] R. Blanco et al. “Entity recommendations in web search”. In: InternationalSemantic Web Conference. Springer. 2013, pp. 33–48.
CIKM 2017 Knowledge Graphs: In Theory and Practice 43/47
![Page 129: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/129.jpg)
References ii
[7] M. Bron, K. Balog, and M. De Rijke. “Ranking related entities: components andanalyses”. In: Proceedings of the 19th ACM international conference onInformation and knowledge management. ACM. 2010, pp. 1079–1088.
[8] R. Das et al. “Chains of reasoning over entities, relations, and text usingrecurrent neural networks”. In: arXiv preprint arXiv:1607.01426 (2016).
[9] A. Fokoue et al. “Predicting drug-drug interactions through large-scalesimilarity-based link prediction”. In: International Semantic Web Conference.Springer. 2016, pp. 774–789.
[10] A. Gattani et al. “Entity extraction, linking, classification, and tagging for socialmedia: a wikipedia-based approach”. In: Proceedings of the VLDB Endowment6.11 (2013), pp. 1126–1137.
[11] S. Guo, M.-W. Chang, and E. Kiciman. “To Link or Not to Link? A Study onEnd-to-End Tweet Entity Linking.”. In: HLT-NAACL. 2013, pp. 1020–1030.
[12] B. Hachey et al. “Evaluating Entity Linking with Wikipedia”. In: Artif. Intell. 194(Jan. 2013), pp. 130–150.
CIKM 2017 Knowledge Graphs: In Theory and Practice 44/47
![Page 130: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/130.jpg)
References iii
[13] J. Hoffart et al. “Robust disambiguation of named entities in text”. In:Proceedings of the Conference on Empirical Methods in Natural LanguageProcessing. Association for Computational Linguistics. 2011, pp. 782–792.
[14] J. Huang et al. “Generating Recommendation Evidence Using TranslationModel.”. In: IJCAI. 2016, pp. 2810–2816.
[15] Y. Lin et al. “Learning Entity and Relation Embeddings for Knowledge GraphCompletion.”. In: AAAI. 2015, pp. 2181–2187.
[16] C. D. Manning et al. “The Stanford CoreNLP Natural Language Processing Toolkit”.In: Association for Computational Linguistics (ACL) System Demonstrations.2014, pp. 55–60.
[17] D. Nadeau and S. Sekine. “A survey of named entity recognition andclassification”. In: Lingvisticae Investigationes 30.1 (2007), pp. 3–26.
[18] M. e. a. Nagarajan. “Predicting Future Scientific Discoveries Based on aNetworked Analysis of the Past Literature”. In: Proceedings of the 21th ACMSIGKDD International Conference on Knowledge Discovery and Data Mining. KDD’15. Sydney, NSW, Australia: ACM, 2015, pp. 2019–2028.
CIKM 2017 Knowledge Graphs: In Theory and Practice 45/47
![Page 131: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/131.jpg)
References iv
[19] J. Pound, P. Mika, and H. Zaragoza. “Ad-hoc Object Retrieval in the Web of Data”.In: Proceedings of the 19th International Conference on World Wide Web. WWW’10. Raleigh, North Carolina, USA: ACM, 2010, pp. 771–780.
[20] L. Ratinov et al. “Local and global algorithms for disambiguation to wikipedia”.In: Proceedings of the 49th Annual Meeting of the Association forComputational Linguistics: Human Language Technologies-Volume 1.Association for Computational Linguistics. 2011, pp. 1375–1384.
[21] M. Schuhmacher, L. Dietz, and S. Paolo Ponzetto. “Ranking entities for webqueries through text and knowledge”. In: Proceedings of the 24th ACMInternational on Conference on Information and Knowledge Management. ACM.2015, pp. 1461–1470.
[22] R. Socher et al. “Reasoning with neural tensor networks for knowledge basecompletion”. In: Advances in neural information processing systems. 2013,pp. 926–934.
[23] F. M. Suchanek, G. Kasneci, and G. Weikum. “Yago: a core of semanticknowledge”. In: Proceedings of the 16th international conference on World WideWeb. ACM. 2007, pp. 697–706.
CIKM 2017 Knowledge Graphs: In Theory and Practice 46/47
![Page 132: Knowledge Graphs: In Theory and Practicesumitbhatia.net/papers/KG_Tutorial_CIKM17_part2.pdf · KnowledgeGraphs:InTheoryandPractice SumitBhatia1andNitishAggarwal2 1IBMResearch,NewDelhi,India](https://reader033.vdocument.in/reader033/viewer/2022050222/5f6758938bc90532dc2b4cc9/html5/thumbnails/132.jpg)
References v
[24] N. Voskarides et al. “Learning to explain entity relationships in knowledgegraphs”. In: Proceedings of the 53rd Annual Meeting of the Association forComputational Linguistics and The 7th International Joint Conference onNatural Language Processing of the Asian Federation of Natural LanguageProcessing (ACL-IJCNLP 2015). 2015, p. 11.
[25] Z. Wang et al. “Knowledge Graph and Text Jointly Embedding.”. In: EMNLP.Vol. 14. 2014, pp. 1591–1601.
[26] I. H. Witten and D. N. Milne. “An effective, low-cost measure of semanticrelatedness obtained from Wikipedia links”. In: (2008).
[27] Y. Zhang, G. Cheng, and Y. Qu. “Towards exploratory relationship search: Aclustering-based approach”. In: Joint International Semantic TechnologyConference. Springer. 2013, pp. 277–293.
[28] H. Zhong et al. “Aligning Knowledge and Text Embeddings by EntityDescriptions.”. In: EMNLP. 2015, pp. 267–272.
CIKM 2017 Knowledge Graphs: In Theory and Practice 47/47