Modeling naturalistic argumentation in research literatures: Representation and interaction design issues

From AcaWiki
Jump to: navigation, search

Citation: Simon J. Buckingham Shum, Victoria Uren, Gangmin Li, Bertrand Sereno, Clara Mancini (2007) Modeling naturalistic argumentation in research literatures: Representation and interaction design issues. International Journal of Intelligent Systems (RSS)
DOI (original publisher): 10.1002/int.20188
Semantic Scholar (metadata): 10.1002/int.20188
Sci-Hub (fulltext): 10.1002/int.20188
Internet Archive Scholar (search for fulltext): Modeling naturalistic argumentation in research literatures: Representation and interaction design issues
Tagged: scholarly publishing (RSS), argumentation (RSS), digital libraries (RSS), interaction design (RSS), computer-supported argumentation (RSS), Scholarly Ontologies project (RSS), ClaiMaker (RSS), ClaiMapper (RSS), ClaimSpotter (RSS), ClaimFinder (RSS)


This paper imagines augmenting scientific communication with "a complementary infrastructure that is “native” to the emerging semantic, collaborative web". It then discusses the Scholarly Ontologies project.

Digital libraries do not answer fundamental questions such as

  • Which publications support and challenge this document?
  • What is the intellectual lineage of this idea?
  • What data are there to support this specific claim or prediction?
  • Who else is working on this problem?
  • Has this approach been used in other fields?
  • What logical or analogical connections have been made between these ideas?

These questions "require semantic annotation at a different level from that addressed by conventional metadata or ontologically based markup in semantic web research, which seek to iron out inconsistency, ambiguity, and incompleteness in the way resources are characterized". In particular, disagreement is important, because this is what marks something as research.

Scholarly Ontologies, Cognitive Coherence Relations, ClaiMaker, ClaiMapper, ClaimFinder

Next they describe the discourse ontology, and its theoretical framing in Cognitive Coherence Relations theory (see e.g. Modelling discourse in contested domains: A semiotic and cognitive framework, or the fundamental linguistics papers Coherence relations in a cognitive theory of discourse representation and Towards a taxonomy of coherence relations).

This provides details about the first ClaiMaker prototype, a ClaiMaker Word plug-in, and the pen-and-paper sketching that helped them understand what was missing. The need for a "big picture" led to ClaiMapper, which allows users to draw links between nodes, and to copy nodes by reference.


One problem with ClaimMapper is that users don't always know how to structure their maps. ClaimSpotter deals with this, in three ways:

  1. By identifying where the author presents and defend arguments (along with ideas for how to break that up into arguments), using text analysis (drawing from CARS, see Summarizing scientific articles: Experiments with relevance and rhetorical status for the method and From scholarly documents to interpretative claims: An approach to bridge the formalisation gulf for the simplification method applied.)

draws from research in hypertext, HCI, CSCW, and computational linguistics.

  1. Users can collaborate, using Claims and Concepts others have added. "This provides a form of extended “semantic cocitation” that exploits the web of structured annotations and extends the citations of a document."
  2. The interface allows "capture/editing/construction of claims" based on automated suggestions

ClaimSpotter interface 2007.png

Formality/Usability Balance

Based on user comments, they "believe that the design of interfaces for creating claim networks, and possibly argument models in general, may influence the kinds and quality of models produced." They argue that good interfaces can help users to "push the balance toward formality".

ClaimFinder and Example Use Cases

ClaimFinder generates visualizations of Concepts/Sets/Claims. Various screenshots are presented, along with two use cases, providing discovery services in a Web browser, with a Java applet or alternately a stand-alone Java application.

They provide examples for two use cases:

  1. Perspective Analysis: “What arguments are there against this paper?”
  2. Lineage Analysis: "Where did this idea come from?"

These motivate what they call "discourse taxonomy" (the annotation process they have been describing).