Completed my thesis under the supervision of Colm O'Riordan over the course of one module Artificial Intelligence Project (CT5129) at 30 ECTS.

My self-proposed thesis is titled "GAPRS: Graph-based Academic Paper Recommendation System" and focussed on creating a graph and network based academic paper recommender system via the OpenAlex API and made use of generative AI to incorporate HCI, XAI and aid interpretation of data visualizations generated by GAPRS which come in the form of graphs and networks.

Deliverables

• Submitted 68-page thesis to examiners for review.

Outputs

• CLI tool which integrates the OpenAlex API for academic papers and ChatGPT for XAI and HCI.

• Market research and competitor analysis of research paper recommender systems (RPRS).

• Historical survey of research paper recommender systems.

• Literature review consisting of the following: identifying consensus in the RPRS space, identifying and summarizing problems facing RPRS community, identifying and collecting open and emerging challenges in the RPRS community, identifying and collecting open challenges in the graph learning recommender systems community, identifying relevant work in related research fields.

• Compilation of the following: List of known recommender system evaluation metrics to date, list of relevant network science metrics, list of known recommendation algorithms to date, list of direct and indirect competitors to GAPRS.

• Novel methodology for approaching problem of recommending academic papers to users via creation of new network science metric, avg(NCCC, BCCC). As well as creation of hybrid weighted undirected 1-Degree and 1.5-Degree multi-egocentric citation networks.

Outcomes

• Awarded Microsoft University Mentorship 2023 which transformed my thesis into an industry-research collaboration.