Investigating transformer-based natural language processing (NLP) techniques and advanced machine learning approaches to improve virtual assistant operations

Conversational artificial intelligence (AI) chatbots, also known as Virtual Assistants (VAs), combine the power of machine learning (ML) and natural language processing (NLP) to understand the intent of a user’s utterance, and to formulate a useful, human-like response. Operating a Virtual Assistant brings many complex challenges; this project strives to address these challenges through undertaking an analytical and empirical assessment of state-of-the-art NLP and deep learning (DL) techniques, with the ultimate goal of enhancing VA-human interactivity to assess a state of wellbeing, overcoming a major hurdle in tailoring wellness management. Interns will develop in-demand skills and expertise in NLP, as well as advanced ML and DL techniques. Through Wysdom, interns will receive access to real-world data and proof-of-concept use cases to test and validate AI product suites. Wysdom will benefit from rigorous research into novel approaches that address the challenges associated with VA operations, leading to improved core business processes.

Faculty Supervisor:

Stephen Thomas;Xiaodan Zhu;Ceren Kolsarici

Student:

Partner:

Wysdom AI

Discipline:

Business

Sector:

Information and cultural industries; Professional, scientific and technical services

University:

Queen's University

Program: