Dr Kulvinder Panesar

Research

This research area is multi-disciplinary spanning AI, data science, agent thinking, linguistics, computational linguistics, NLP, KR, and the Semantic Web.  She has conceptually designed and developed  a linguistically text based conversational software agent (LING-CSA) framework, addressing the integration, intersection, and interface of language, knowledge, and speech act constructions (SAC).  LING-CSA is a Java based prototype developed in Eclipse. 

Her research contributions included: (i) extending the theoretical and computational adequacy of the linguistic theory - Role and Reference Grammar (RRG); (ii) integrating the RRG language model  with concept of speech act constructions (SAC) as the linguistic engine; (iii) motivating an agent framework  intersecting with the  linguistic engine,  an agent cognitive and dialogue model to facilitate conversation implemented as a proof-of-concept; (4) insights into the language/knowledge representation interface.

Current research interest involves demystifying statistical vs lingustic NLP for conversational software agents, with the goal to investigating linguistic NLP with statistical  NLP learning and support – for s  forward-thinking  hybrid approach to conversational AI.  To explore and apply this research the healthcare domain is selected. Here the focus on language and cognition in mild cognitive impairment and patients living with dementia is to be explored based on a   language phenomenon  and further explored for  automation as an intervention embeded into a  conversational agent with automatic speech recognition.