SlideShare a Scribd company logo
1 of 25
Natural Intelligence -
Commonsense Question Answering
with Conceptual Graphs

  Fatih Mehmet Güler and Aysenur Birturk
  Department of Computer Engineering, METU
          06531, Ankara/TURKEY
             fmguler@gmail.com
          birturk@ceng.metu.edu.tr
Motivation

   Massive Knowledge found as Natural Language
   Text based Question Answering (no tagging)
   Open Domain Question Answering
   Address Commonsense Reasoning Problem
   Linguistically motivated KRR
       Intelligence is the accumulation of knowledge
   Integrate State of the Art Tools
   Ultimate goal: Getting closer to strong AI
Summary of the System

   Natural Language is parsed
   Utterances are represented using CGs
   Concepts and Relation types are mapped to
    Cyc equivalent counterparts
   Type hierarchies are computed
   Knowledge is accumulated
   If the input is a question
       Search for answer (projection)
Summary of the System (Cont’d)

                       NI




           NLP        KRR       Commonsense




           CCG        CGs        Open Cyc




         C&C Tools   Cogitant
Background

   Combinatory Categorial Grammar (CCG)
       C&C Tools
   Conceptual Graphs
       Cogitant
   Open Cyc
Combinatory Categorial Grammar (CCG)

   Lexicalized Theory of Grammar based on
    Categorial Grammar ( Steedman 2001).
       Functions can be applied or composed
       Arguments can be picked up or turned into
        functors (Type raising)
   Easy for Semantic Representations
       Small number of semantically transparent
        combinatory rules to combine CCG categories.
           Assign semantic representations to the lexical entries
           Interpret combinatory rules
CCG parse for “Mr. Hyde ate two
lemmons”
CCG Parse for “Susan knows that Bob
likes Fred”
DRS for “Susan knows that Bob likes
Fred”
C&C Tools
   Linguistically Motivated Large-Scale NLP with C&C
    and Boxer. (Curran, Clark, Bos, 2007)
       C&C Parser
         POS Tagging, Supertagging

         Parsing, Chunking

         Named Entity Recognition

       Boxer
         Uses CCG parser output

         Generates DRS Semantic Representations

   Freely available for research
       http://svn.ask.it.usyd.edu.au/trac/candc/wiki
C&C Tools

   Large Scale NLP is possible with C&C and
    Boxer
   C&C Parser: state of the art parser for CCG
   Boxer: Semantic representations in DRS
Open Cyc

   Open source version of Cyc system
   Cyc: greatest effort to encode Common Sense
    knowledge in machine processable way
   500.000 concepts 26.000 relations and 5.000.000
    assertions
   CycL language similar to Lisp
   We use Cyc to map parsed words to common sense
    counterparts such as person to #$Person
    (disambiguation)
Open Cyc (cont’d)

   (#$likesAsFriend #$GeorgeWBush #$AlGore)
   #$isa, #$genls
   (#$isa #$GeorgeWBush
    #$UnitedStatesPresident)
   (#$genls #$UnitedStatesPresident #$Person)
Cogitant

   Library for Conceptual Graph operations
   Supports broad CG operations (Genest &
    Salvat, 1998)
       Graph representation
       Conversion from CGIF
       Projection checking
       Rule application
Natural Intelligence – Commonsense
Question Answering with CGs
   Augment Common Sense knowledge
   Modular Approach
       Separation of Concerns
       State of the art tools
Architecture - Modules

   Natural Language Processing (C&C Tools are used
    for implementation)
       Convert natural language to CGIF
   Reasoning (Cogitant library is used for
    implementation)
       CG operations
   Common Sense (Open Cyc is used for
    implementation)
       Common sense mapping
   Storage (Conceptual Graphs are stored in a
    database)
       Persistence of CGs
System Definition
   User enters a sentence from web interface;
   This sentence is converted to CGIF using the NLP module;
   CGIF is converted to CGs using the reasoning module;
   Support is generated to CGs using the common sense module;
   Common sense rules gathered from common sense module are
    applied to CGs using reasoning module;
   CGs are merged to the previous ones using reasoning module;
   If the input sentence is a question sentence, same operations
    take place, except the resulting graph is used to query existing
    CGs using the reasoning service, and if there are projections
    from this query graph to previous CGs, results are displayed to
    the user;
   CGs are persisted using the storage module.
Common Sense Mapping

   Cyc: (prettyString TERM STRING)
   Chain up to #$Thing using #$genls relations
   Same for relations using #$genlPreds
   Relation hierarchies are converted to forward
    rules
       #$performedBy -> #$temporallyRelated
Sample Concept Hierarchy
   #$Place ->
           #$EnduringThing-Localized ->
                        #$Location-Underspecified ->
                                        #$Thing ->
                        #$SomethingExisting ->
                                        #$Individual ->
                                                        #$Thing ^^
                                                        #$Trajector-Underspecified ->
                                                                        #$Location-Underspecified ^^
                                        #$TemporallyExistingThing ->
                                                        #$TemporalThing ->
                                                                        #$Individual ^^
                        #$SpatialThing-Localized ->
                                        #$TemporallyExistingThing ^^
                                        #$SpatialThing ->
                                                        #$Individual ^^
                                        #$Boundary-Underspecified ->
                                                        #$Region-Underspecified ->
                                                                        #$Location-Underspecified ^^
                                                        #$Landmark-Underspecified ->
                                                                        #$Individual ^^
                                                                        #$Location-Underspecified ^^
                        #$SpatialThing-NonSituational ->
                                        #$SpatialThing ^^
                                        #$Individual ^^
           #$Location-Underspecified ^^
Conversion to Cogitant Support

   Convert Cyc hierarcy to Cogitant support
    format
       Concept Types
       Relation Types
       Individuals
       Rules
   Convert assertions to Cogitant graph format
   Apply forward rules
Answering Queries
Significance

   Sentences like;
       What are the intangible things in this situation?
       Was Mr. Hyde there while eating the apples?
       Does Mr. Hyde exist after eating the apples?
       Do the apples exist after Mr. Hyde ate them?
   Deep Natural Language Understanding
   State of the art tools
   Open domain question answering
Difficulties

   Open Cyc API is broken
       Does not work in Turkish locale (fixes are sent to
        maintainers)
       Still, provided API sends one IP packet per character, way
        too slow over network
       Custom socket API is developed and used over TCP
       Custom Lisp functions for generalization hierarchy and
        concept mapping
   Cogitant problematic
       Java API is very limited (compared to C++)
       Only works over XML files
Conclusion

   Central Integrated Common Sense QAS
   CCG for Natural Language Processing
   Conceptual Graphs for KRR
   Cyc for Common Sense
Future Work

   Implement Rule Induction
   Backward Chaining (Resolution)
   Improve NLP module and Common Sense mapping
   Probabilistic Reasoning
   Question Answering System (QAS) to be used in;
       Education (Learning Management Systems)
       Semantic Search (Content Management Systems)
       Intelligent Help

More Related Content

Similar to Natural Intelligence - Commonsense Question Answering with Conceptual Graphs

Embrace NoSQL and Eventual Consistency with Ripple
Embrace NoSQL and Eventual Consistency with RippleEmbrace NoSQL and Eventual Consistency with Ripple
Embrace NoSQL and Eventual Consistency with RippleSean Cribbs
 
Overview of running R in the Oracle Database
Overview of running R in the Oracle DatabaseOverview of running R in the Oracle Database
Overview of running R in the Oracle DatabaseBrendan Tierney
 
Leveraging Lucene/Solr as a Knowledge Graph and Intent Engine: Presented by T...
Leveraging Lucene/Solr as a Knowledge Graph and Intent Engine: Presented by T...Leveraging Lucene/Solr as a Knowledge Graph and Intent Engine: Presented by T...
Leveraging Lucene/Solr as a Knowledge Graph and Intent Engine: Presented by T...Lucidworks
 
PHP Development With MongoDB
PHP Development With MongoDBPHP Development With MongoDB
PHP Development With MongoDBFitz Agard
 
PHP Development with MongoDB (Fitz Agard)
PHP Development with MongoDB (Fitz Agard)PHP Development with MongoDB (Fitz Agard)
PHP Development with MongoDB (Fitz Agard)MongoSF
 
HPTS talk on micro-sharding with Katta
HPTS talk on micro-sharding with KattaHPTS talk on micro-sharding with Katta
HPTS talk on micro-sharding with KattaTed Dunning
 
Mining Software Archives to Support Software Development
Mining Software Archives to Support Software DevelopmentMining Software Archives to Support Software Development
Mining Software Archives to Support Software DevelopmentThomas Zimmermann
 
[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative Filtering[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative FilteringYONG ZHENG
 
Building DSLs On CLR and DLR (Microsoft.NET)
Building DSLs On CLR and DLR (Microsoft.NET)Building DSLs On CLR and DLR (Microsoft.NET)
Building DSLs On CLR and DLR (Microsoft.NET)Vitaly Baum
 
Code as Data workshop: Using source{d} Engine to extract insights from git re...
Code as Data workshop: Using source{d} Engine to extract insights from git re...Code as Data workshop: Using source{d} Engine to extract insights from git re...
Code as Data workshop: Using source{d} Engine to extract insights from git re...source{d}
 
Stuff About CQRS
Stuff About CQRSStuff About CQRS
Stuff About CQRSthinkddd
 
Compass Framework
Compass FrameworkCompass Framework
Compass FrameworkLukas Vlcek
 
R Workshop for Beginners
R Workshop for BeginnersR Workshop for Beginners
R Workshop for BeginnersMetamarkets
 
Concurrent and Distributed Applications with Akka, Java and Scala
Concurrent and Distributed Applications with Akka, Java and ScalaConcurrent and Distributed Applications with Akka, Java and Scala
Concurrent and Distributed Applications with Akka, Java and ScalaFernando Rodriguez
 
High Performance Ruby: Evented vs. Threaded
High Performance Ruby: Evented vs. ThreadedHigh Performance Ruby: Evented vs. Threaded
High Performance Ruby: Evented vs. ThreadedEngine Yard
 
Introduction into scalable graph analysis with Apache Giraph and Spark GraphX
Introduction into scalable graph analysis with Apache Giraph and Spark GraphXIntroduction into scalable graph analysis with Apache Giraph and Spark GraphX
Introduction into scalable graph analysis with Apache Giraph and Spark GraphXrhatr
 
Tips And Tricks For Bioinformatics Software Engineering
Tips And Tricks For Bioinformatics Software EngineeringTips And Tricks For Bioinformatics Software Engineering
Tips And Tricks For Bioinformatics Software Engineeringjtdudley
 

Similar to Natural Intelligence - Commonsense Question Answering with Conceptual Graphs (20)

Embrace NoSQL and Eventual Consistency with Ripple
Embrace NoSQL and Eventual Consistency with RippleEmbrace NoSQL and Eventual Consistency with Ripple
Embrace NoSQL and Eventual Consistency with Ripple
 
Spatial LDA
Spatial LDASpatial LDA
Spatial LDA
 
Overview of running R in the Oracle Database
Overview of running R in the Oracle DatabaseOverview of running R in the Oracle Database
Overview of running R in the Oracle Database
 
Introduction to Domain-Driven Design
Introduction to Domain-Driven DesignIntroduction to Domain-Driven Design
Introduction to Domain-Driven Design
 
Leveraging Lucene/Solr as a Knowledge Graph and Intent Engine: Presented by T...
Leveraging Lucene/Solr as a Knowledge Graph and Intent Engine: Presented by T...Leveraging Lucene/Solr as a Knowledge Graph and Intent Engine: Presented by T...
Leveraging Lucene/Solr as a Knowledge Graph and Intent Engine: Presented by T...
 
PHP Development With MongoDB
PHP Development With MongoDBPHP Development With MongoDB
PHP Development With MongoDB
 
PHP Development with MongoDB (Fitz Agard)
PHP Development with MongoDB (Fitz Agard)PHP Development with MongoDB (Fitz Agard)
PHP Development with MongoDB (Fitz Agard)
 
HPTS talk on micro-sharding with Katta
HPTS talk on micro-sharding with KattaHPTS talk on micro-sharding with Katta
HPTS talk on micro-sharding with Katta
 
Mining Software Archives to Support Software Development
Mining Software Archives to Support Software DevelopmentMining Software Archives to Support Software Development
Mining Software Archives to Support Software Development
 
[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative Filtering[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative Filtering
 
Building DSLs On CLR and DLR (Microsoft.NET)
Building DSLs On CLR and DLR (Microsoft.NET)Building DSLs On CLR and DLR (Microsoft.NET)
Building DSLs On CLR and DLR (Microsoft.NET)
 
Code as Data workshop: Using source{d} Engine to extract insights from git re...
Code as Data workshop: Using source{d} Engine to extract insights from git re...Code as Data workshop: Using source{d} Engine to extract insights from git re...
Code as Data workshop: Using source{d} Engine to extract insights from git re...
 
Stuff About CQRS
Stuff About CQRSStuff About CQRS
Stuff About CQRS
 
Compass Framework
Compass FrameworkCompass Framework
Compass Framework
 
R Workshop for Beginners
R Workshop for BeginnersR Workshop for Beginners
R Workshop for Beginners
 
Concurrent and Distributed Applications with Akka, Java and Scala
Concurrent and Distributed Applications with Akka, Java and ScalaConcurrent and Distributed Applications with Akka, Java and Scala
Concurrent and Distributed Applications with Akka, Java and Scala
 
High Performance Ruby: Evented vs. Threaded
High Performance Ruby: Evented vs. ThreadedHigh Performance Ruby: Evented vs. Threaded
High Performance Ruby: Evented vs. Threaded
 
Jquery2012 defs
Jquery2012 defsJquery2012 defs
Jquery2012 defs
 
Introduction into scalable graph analysis with Apache Giraph and Spark GraphX
Introduction into scalable graph analysis with Apache Giraph and Spark GraphXIntroduction into scalable graph analysis with Apache Giraph and Spark GraphX
Introduction into scalable graph analysis with Apache Giraph and Spark GraphX
 
Tips And Tricks For Bioinformatics Software Engineering
Tips And Tricks For Bioinformatics Software EngineeringTips And Tricks For Bioinformatics Software Engineering
Tips And Tricks For Bioinformatics Software Engineering
 

Recently uploaded

Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Farhan Tariq
 
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxDigital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxLoriGlavin3
 
2024 April Patch Tuesday
2024 April Patch Tuesday2024 April Patch Tuesday
2024 April Patch TuesdayIvanti
 
So einfach geht modernes Roaming fuer Notes und Nomad.pdf
So einfach geht modernes Roaming fuer Notes und Nomad.pdfSo einfach geht modernes Roaming fuer Notes und Nomad.pdf
So einfach geht modernes Roaming fuer Notes und Nomad.pdfpanagenda
 
Moving Beyond Passwords: FIDO Paris Seminar.pdf
Moving Beyond Passwords: FIDO Paris Seminar.pdfMoving Beyond Passwords: FIDO Paris Seminar.pdf
Moving Beyond Passwords: FIDO Paris Seminar.pdfLoriGlavin3
 
Transcript: New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024Transcript: New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024BookNet Canada
 
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...panagenda
 
Testing tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesTesting tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesKari Kakkonen
 
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...Wes McKinney
 
UiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPathCommunity
 
The Fit for Passkeys for Employee and Consumer Sign-ins: FIDO Paris Seminar.pptx
The Fit for Passkeys for Employee and Consumer Sign-ins: FIDO Paris Seminar.pptxThe Fit for Passkeys for Employee and Consumer Sign-ins: FIDO Paris Seminar.pptx
The Fit for Passkeys for Employee and Consumer Sign-ins: FIDO Paris Seminar.pptxLoriGlavin3
 
How to write a Business Continuity Plan
How to write a Business Continuity PlanHow to write a Business Continuity Plan
How to write a Business Continuity PlanDatabarracks
 
What is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfWhat is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfMounikaPolabathina
 
Generative AI for Technical Writer or Information Developers
Generative AI for Technical Writer or Information DevelopersGenerative AI for Technical Writer or Information Developers
Generative AI for Technical Writer or Information DevelopersRaghuram Pandurangan
 
Decarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityDecarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityIES VE
 
Emixa Mendix Meetup 11 April 2024 about Mendix Native development
Emixa Mendix Meetup 11 April 2024 about Mendix Native developmentEmixa Mendix Meetup 11 April 2024 about Mendix Native development
Emixa Mendix Meetup 11 April 2024 about Mendix Native developmentPim van der Noll
 
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptx
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptxPasskey Providers and Enabling Portability: FIDO Paris Seminar.pptx
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptxLoriGlavin3
 
From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .Alan Dix
 
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024BookNet Canada
 
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptxThe Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptxLoriGlavin3
 

Recently uploaded (20)

Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...Genislab builds better products and faster go-to-market with Lean project man...
Genislab builds better products and faster go-to-market with Lean project man...
 
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxDigital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
 
2024 April Patch Tuesday
2024 April Patch Tuesday2024 April Patch Tuesday
2024 April Patch Tuesday
 
So einfach geht modernes Roaming fuer Notes und Nomad.pdf
So einfach geht modernes Roaming fuer Notes und Nomad.pdfSo einfach geht modernes Roaming fuer Notes und Nomad.pdf
So einfach geht modernes Roaming fuer Notes und Nomad.pdf
 
Moving Beyond Passwords: FIDO Paris Seminar.pdf
Moving Beyond Passwords: FIDO Paris Seminar.pdfMoving Beyond Passwords: FIDO Paris Seminar.pdf
Moving Beyond Passwords: FIDO Paris Seminar.pdf
 
Transcript: New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024Transcript: New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
 
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
Why device, WIFI, and ISP insights are crucial to supporting remote Microsoft...
 
Testing tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesTesting tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examples
 
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
The Future Roadmap for the Composable Data Stack - Wes McKinney - Data Counci...
 
UiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to Hero
 
The Fit for Passkeys for Employee and Consumer Sign-ins: FIDO Paris Seminar.pptx
The Fit for Passkeys for Employee and Consumer Sign-ins: FIDO Paris Seminar.pptxThe Fit for Passkeys for Employee and Consumer Sign-ins: FIDO Paris Seminar.pptx
The Fit for Passkeys for Employee and Consumer Sign-ins: FIDO Paris Seminar.pptx
 
How to write a Business Continuity Plan
How to write a Business Continuity PlanHow to write a Business Continuity Plan
How to write a Business Continuity Plan
 
What is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfWhat is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdf
 
Generative AI for Technical Writer or Information Developers
Generative AI for Technical Writer or Information DevelopersGenerative AI for Technical Writer or Information Developers
Generative AI for Technical Writer or Information Developers
 
Decarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityDecarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a reality
 
Emixa Mendix Meetup 11 April 2024 about Mendix Native development
Emixa Mendix Meetup 11 April 2024 about Mendix Native developmentEmixa Mendix Meetup 11 April 2024 about Mendix Native development
Emixa Mendix Meetup 11 April 2024 about Mendix Native development
 
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptx
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptxPasskey Providers and Enabling Portability: FIDO Paris Seminar.pptx
Passkey Providers and Enabling Portability: FIDO Paris Seminar.pptx
 
From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .
 
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
 
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptxThe Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
 

Natural Intelligence - Commonsense Question Answering with Conceptual Graphs

  • 1. Natural Intelligence - Commonsense Question Answering with Conceptual Graphs Fatih Mehmet Güler and Aysenur Birturk Department of Computer Engineering, METU 06531, Ankara/TURKEY fmguler@gmail.com birturk@ceng.metu.edu.tr
  • 2. Motivation  Massive Knowledge found as Natural Language  Text based Question Answering (no tagging)  Open Domain Question Answering  Address Commonsense Reasoning Problem  Linguistically motivated KRR  Intelligence is the accumulation of knowledge  Integrate State of the Art Tools  Ultimate goal: Getting closer to strong AI
  • 3. Summary of the System  Natural Language is parsed  Utterances are represented using CGs  Concepts and Relation types are mapped to Cyc equivalent counterparts  Type hierarchies are computed  Knowledge is accumulated  If the input is a question  Search for answer (projection)
  • 4. Summary of the System (Cont’d) NI NLP KRR Commonsense CCG CGs Open Cyc C&C Tools Cogitant
  • 5. Background  Combinatory Categorial Grammar (CCG)  C&C Tools  Conceptual Graphs  Cogitant  Open Cyc
  • 6. Combinatory Categorial Grammar (CCG)  Lexicalized Theory of Grammar based on Categorial Grammar ( Steedman 2001).  Functions can be applied or composed  Arguments can be picked up or turned into functors (Type raising)  Easy for Semantic Representations  Small number of semantically transparent combinatory rules to combine CCG categories.  Assign semantic representations to the lexical entries  Interpret combinatory rules
  • 7. CCG parse for “Mr. Hyde ate two lemmons”
  • 8. CCG Parse for “Susan knows that Bob likes Fred”
  • 9. DRS for “Susan knows that Bob likes Fred”
  • 10. C&C Tools  Linguistically Motivated Large-Scale NLP with C&C and Boxer. (Curran, Clark, Bos, 2007)  C&C Parser  POS Tagging, Supertagging  Parsing, Chunking  Named Entity Recognition  Boxer  Uses CCG parser output  Generates DRS Semantic Representations  Freely available for research  http://svn.ask.it.usyd.edu.au/trac/candc/wiki
  • 11. C&C Tools  Large Scale NLP is possible with C&C and Boxer  C&C Parser: state of the art parser for CCG  Boxer: Semantic representations in DRS
  • 12. Open Cyc  Open source version of Cyc system  Cyc: greatest effort to encode Common Sense knowledge in machine processable way  500.000 concepts 26.000 relations and 5.000.000 assertions  CycL language similar to Lisp  We use Cyc to map parsed words to common sense counterparts such as person to #$Person (disambiguation)
  • 13. Open Cyc (cont’d)  (#$likesAsFriend #$GeorgeWBush #$AlGore)  #$isa, #$genls  (#$isa #$GeorgeWBush #$UnitedStatesPresident)  (#$genls #$UnitedStatesPresident #$Person)
  • 14. Cogitant  Library for Conceptual Graph operations  Supports broad CG operations (Genest & Salvat, 1998)  Graph representation  Conversion from CGIF  Projection checking  Rule application
  • 15. Natural Intelligence – Commonsense Question Answering with CGs  Augment Common Sense knowledge  Modular Approach  Separation of Concerns  State of the art tools
  • 16. Architecture - Modules  Natural Language Processing (C&C Tools are used for implementation)  Convert natural language to CGIF  Reasoning (Cogitant library is used for implementation)  CG operations  Common Sense (Open Cyc is used for implementation)  Common sense mapping  Storage (Conceptual Graphs are stored in a database)  Persistence of CGs
  • 17. System Definition  User enters a sentence from web interface;  This sentence is converted to CGIF using the NLP module;  CGIF is converted to CGs using the reasoning module;  Support is generated to CGs using the common sense module;  Common sense rules gathered from common sense module are applied to CGs using reasoning module;  CGs are merged to the previous ones using reasoning module;  If the input sentence is a question sentence, same operations take place, except the resulting graph is used to query existing CGs using the reasoning service, and if there are projections from this query graph to previous CGs, results are displayed to the user;  CGs are persisted using the storage module.
  • 18. Common Sense Mapping  Cyc: (prettyString TERM STRING)  Chain up to #$Thing using #$genls relations  Same for relations using #$genlPreds  Relation hierarchies are converted to forward rules  #$performedBy -> #$temporallyRelated
  • 19. Sample Concept Hierarchy  #$Place ->  #$EnduringThing-Localized ->  #$Location-Underspecified ->  #$Thing ->  #$SomethingExisting ->  #$Individual ->  #$Thing ^^  #$Trajector-Underspecified ->  #$Location-Underspecified ^^  #$TemporallyExistingThing ->  #$TemporalThing ->  #$Individual ^^  #$SpatialThing-Localized ->  #$TemporallyExistingThing ^^  #$SpatialThing ->  #$Individual ^^  #$Boundary-Underspecified ->  #$Region-Underspecified ->  #$Location-Underspecified ^^  #$Landmark-Underspecified ->  #$Individual ^^  #$Location-Underspecified ^^  #$SpatialThing-NonSituational ->  #$SpatialThing ^^  #$Individual ^^  #$Location-Underspecified ^^
  • 20. Conversion to Cogitant Support  Convert Cyc hierarcy to Cogitant support format  Concept Types  Relation Types  Individuals  Rules  Convert assertions to Cogitant graph format  Apply forward rules
  • 22. Significance  Sentences like;  What are the intangible things in this situation?  Was Mr. Hyde there while eating the apples?  Does Mr. Hyde exist after eating the apples?  Do the apples exist after Mr. Hyde ate them?  Deep Natural Language Understanding  State of the art tools  Open domain question answering
  • 23. Difficulties  Open Cyc API is broken  Does not work in Turkish locale (fixes are sent to maintainers)  Still, provided API sends one IP packet per character, way too slow over network  Custom socket API is developed and used over TCP  Custom Lisp functions for generalization hierarchy and concept mapping  Cogitant problematic  Java API is very limited (compared to C++)  Only works over XML files
  • 24. Conclusion  Central Integrated Common Sense QAS  CCG for Natural Language Processing  Conceptual Graphs for KRR  Cyc for Common Sense
  • 25. Future Work  Implement Rule Induction  Backward Chaining (Resolution)  Improve NLP module and Common Sense mapping  Probabilistic Reasoning  Question Answering System (QAS) to be used in;  Education (Learning Management Systems)  Semantic Search (Content Management Systems)  Intelligent Help