Creative Computing in Language Understanding: A Novel Approach to Pragmatic Analysis
This thesis introduces a Creative Computing-based Pragmatics Analysis Approach crafted to help computational systems understand human language subtleties in English text. It is driven by the need for methodologies that go beyond parsing linguistic structures to include the implicit meanings and intentions in humans. A primary contribution of this research is the development of the Creative Language Analysing Tool, a prototype that embodies the integration of syntax, semantics, and pragmatics into a unified model. This integration provides an in-depth view of language understanding, presenting it as a complex entity that serves as a dynamic means of communication and interaction.
The research explores pragmatics analysis, highlighting its importance in linguistics and the benefits of combining it with computing. The Data-Information-Knowledge-Wisdom (DIKW) pyramid is employed as a foundational framework, focusing on its representation layers, transformation processes, and their relation to language understanding. By examining linguistic levels, speech acts, intentions, and implicature, the thesis integrates the DIKW model with syntax, semantics, and pragmatics analysis.
The research is corroborated through experiments demonstrating the functionality of the Creative Language Analysing Tool - a language understanding prototype, highlighting its effectiveness in conducting syntax, semantics, and pragmatics analysis. This thesis concludes by synthesising the outcomes of this interdisciplinary study, evaluating the approach’s efficacy in pragmatics analysis, and illustrating its significant contributions to the realms of creative computing and language understanding. This research not only analyses language to uncover hidden meanings within inputs but also substantiates its findings through literature validation, leveraging the capabilities of prototypes to underscore the potential for ongoing advancements in this interdisciplinary field. The successful deployment of this prototype and its novel analytical framework propels the discipline forward and sets a new benchmark
History
Supervisor(s)
Hongji YangDate of award
2024-06-24Author affiliation
School of Computing and Mathematical SciencesAwarding institution
University of LeicesterQualification level
- Doctoral
Qualification name
- PhD