Using cDnS as an upper-level framework for a Scholarly Debate Ontology (FOIS 2010 talk)

Presentation of paper accepted at the the Sixth International Conference on Formal Ontology in Information Systems (FOIS 2010), Toronto, 12--14 May 2010.

Transcript

1. Using cDnS as an upper-level framework for a Scholarly Debate Ontology FOIS 2010, 12—14 May, 2010 Neil BENN Knowledge Media Institute, The Open University
2. Library & IS problem •  Need to move beyond accessing academic documents – search engines, digital libraries, e-journals, e-prints, etc. •  Need support for analysing academic domains to determine (e.g.) – Who are the experts? – What are the canonical papers? – What is the leading edge?
3. Example Analysis
4. Example Analysis
5. Research Focus •  Analysing academic domains  understanding scholarly debate •  Mapping and analysis of debate in academic domains – What is the structure of the ongoing dialogue? – What are the controversial issues? – What are the main bodies of opinion?
6. Analysing scholarly debate •  Structural features of the debate not easily detected from raw source material •  E.g. Detecting clusters of viewpoints in the debate – Used to supplement (not replace) source material
7. Example Analysis Visualisation and clustering performed using NetDraw BASIC-ANTI-ABORTION-ARGUMENT BASIC-PRO-ABORTION-ARGUMENT BODILY-RIGHTS-ARGUMENT ABORTION-BREAST-CANCER-HYPOTHESIS TACIT-CONSENT-OBJECTION-ARGUMENT EQUALITY-OBJECTION-ARGUMENT CONTRACEPTION-OBJECTION-ARGUMENT RESPONSIBILITY-OBJECTION-ARGUMENT JUDITH_THOMSON DON_MARQUIS PETER_SINGER ERIC_OLSON DEAN_STRETTON MICHAEL_TOOLEY
8. Scholarly Debate Ontology •  Based on Robert Horn debate mapping approach, theorised as ‘logic of debate’ in Yoshimi (2004) •  Representing dialectical exchange between scholars Concerned with macro-argument structure
9. Scholarly Debate Ontology •  Essential elements for representing debate in academic domains Minimal commitment to start with •  References upper-level constructivist Descriptions and Situations (cDnS) ontology Motivated by view of academic domains as settings for collective construction of knowledge
10. cDnS ontology •  Expressivity to talk about social contexts in which collectives make and produce sense Ontology of collective sensemaking or collective knowledge construction •  Core “information object” configuration of cDnS
11. cdns:InformationObject •  Vehicles for communicating content between agents •  Expression / Sign-vehicle •  Needs physical realisation One IO can have multiple realisations •  Status of different modes of expression? Is orally-delivered speech the same expression as written speech but just different realisation? Or are they two different expressions?
12. sdo:Publication cdns:InformationObject sdo:Publication •  Academic Publications main vehicles for communicating scholarly knowledge •  IO = Publication as a whole •  IO = Parts of publication Verbal expressions: clauses, sentences Hybrid (verbal + non-verbal) expressions: tables, graphs, figures
13. cdns:Description
14. cdns:Description •  The content expressed by an Information Object •  Interpretant •  Same content can be expressed by different InformationObjects
15. cdns:Description •  Thesis (Description) expressed by Academic Publication (Information Object) cf. Plot expressed by novel •  Propositional content expressed by clauses and sentences •  Non-propositional content expressed by questions (interrogative sentences) But have seen elsewhere questions characterised as propositional
16. sdo:Issue •  Context for argumentative exchange •  Typically correspond to Research Questions expressed in academic publications cdns:Description sdo:NonPropositionalContent sdo:Issue
17. sdo:Proposition, sdo:Argument cdns:Description sdo:PropositionalContent sdo:Proposition sdo:Argument •  “supports” & “disputes” argument relations •  “addresses” relation between arguments and issues
18. cdns:Concept
19. cdns:Concept •  Defined by and used by descriptions •  Classify other entities •  Constitution article defines President classifies Barack Obama •  President classifies (at different times) Bill Clinton and George Bush •  Congressman classifies a number of persons at same time
20. sdo:DomainConcept cdns:Concept sdo:DomainConcept •  Elements of specialised vocabulary of an academic domain •  Defined and used by theories •  But what is status of phlogiston as a concept? Defined by a defunct scientific theory Still part of the lexicon But is it part of the conceptual system?
21. cdns:SocialAgent •  Interprets InformationObjects •  Agent conceives the Description expressed by an InformationObject
22. cdns:SocialAgent
23. cdns:SocialAgent •  Two agents (e.g. author and reader) can interpret an Information Object differently thus conceiving different descriptions •  Different descriptions can be contradictory (though ostensibly from the same Information Object) cdns:SocialAgent sdo:Person sdo:Organisation
24. cdns:Collection
25. cdns:Collection •  Has at least two entities as its members •  Emerges out of its member entities (but member entities retain own identity) •  Different from mathematical sets Members can change or be substituted without affecting identity of collection No empty or singleton collections
26. cdns:Collection •  Community of researchers (Knowledge Community) •  Share one or more descriptions that unify collection
27. sdo:Position cdns:Collection sdo:Position •  A Collection of mutually complementary arguments •  Positions can “oppose” each other •  E.g Physicalism v Dualism
28. cdns:Situation
29. cdns:Situation •  A setting for other entities (including other situations) •  Observable state of affairs (requires an agent to conceive it) •  Use of this notion of Situations needs to be explore in future work
30. SDO mapped to cDnS
31. Building up a Debate Network
32. Building up a Debate Network
33. Cluster Analysis Visualisation and clustering performed using NetDraw BASIC-ANTI-ABORTION-ARGUMENT BASIC-PRO-ABORTION-ARGUMENT BODILY-RIGHTS-ARGUMENT ABORTION-BREAST-CANCER-HYPOTHESIS TACIT-CONSENT-OBJECTION-ARGUMENT EQUALITY-OBJECTION-ARGUMENT CONTRACEPTION-OBJECTION-ARGUMENT RESPONSIBILITY-OBJECTION-ARGUMENT JUDITH_THOMSON DON_MARQUIS PETER_SINGER ERIC_OLSON DEAN_STRETTON MICHAEL_TOOLEY
34. Concluding Remarks •  Ontology for representing and reasoning about scholarly debate in academic domains •  Use of cDnS as an upper-level framework for ensuring design captured essential elements of aspect of reality dealing with collective construction of knowledge Does it buy be anything else?
35. Concluding Remarks •  Distributed modelling of debate in academic domains But need to support competing representations of same paper Layering of cdns:Situations? •  Make better use of cDnS axiomatisation of collectives Other relations between Positions besides “opposes”
36. Thank you