|dc.identifier.citation||Computing Arts 2001 : digital resources for research in the humanities : 26th-28th September 2001, Veterinary Science Conference Centre, the University of Sydney / hosted by the Scholarly Text and Imaging Service (SETIS), the University of Sydney Library, and the Research Institute for Humanities and Social Sciences (RIHSS), the University of Sydney||en_AU|
|dc.description.abstract||This paper purports to elaborate and address several issues which lie at the intersection of computational linguistics and psychology. The first issue addressed is that of the interaction between discourse and semantics by virtue of empirical linguistic and psychotherapeutic evidence. This paper then gives a formal account of the knowledge representation and reasoning processes involved in the construction of an XML knowledge base for use in the sematic analysis of psychotherapeutic transcripts. Computational methods for the automatic mark-up and inference of the psychotherapeutic phenomena under investigation are detailed in order to further develop intuitions behind a particular pragmatic theory of language known as the Metamodel.
The work presented here ultimately aims to produce a sustainable system for the evaluation of the effectiveness of any given psychotherapeutic technique. The possibility exists for such a system to recognise successful therapeutic mechanisms and further still, to infer new ones, or suggest improvements, or offer novel explanations as to the success or failure of the therapy itself.
The work discussed here stems from research in computational linguistics, psychotherapy, and philosophy. The corpus used is a culmination of client transcripts taken before, during, and after therapy. The particular therapeutic technique used here is known as the Metamodel (Bandler and Grinder, 1975). The Metamodel was originally proffered as a method of language analysis suitable for use by practitioners of any psychotherapeutic technique. It theorises that speech utterances are related to a clients deep structure through three primary mechanisms, namely generalisation, deletion, and distortion. Previous hand tagging of our data has proven support for such claims. It is our aim to automate the identification and reasoning process. The issues and processes involved in the automation of such tagging are discussed here. Architectural and philosophical issues relating syntax (or grammar), semantics (Larson and Segal, 1995), and pragmatics (Grice, 1989; Searle, 1969) are raised. Discourse Representation Theory (Kamp, 1981; Asher and Lascarides, 1995) is discussed and used here in order to infer discourse relations.||en_AU|
|dc.description.sponsorship||Hosted by the Scholarly Text and Imaging Service (SETIS), the University of Sydney Library, and the Research Institute for Humanities and Social Sciences (RIHSS), the University of Sydney.||en_AU|
|dc.publisher||Research Institute for Humanities and Social Sciences (RIHSS), the University of Sydney.||en_AU|
|dc.rights||Copyright the University of Sydney||en|
|dc.title||Discourse Semantics for the Analysis of Change in Language||en_AU|
|Appears in Collections:||Computing Arts 2001: Digital Resources for Research in the Humanities|