Jinmo Kim’s Preliminary Exam
PhD student Jinmo Kim will present his dissertation proposal, "Resolving Ambiguity, Revealing Dynamics: NLP Approaches to Funding Analysis in Library and Information Science." His preliminary examination committee includes Associate Professor Jana Diesner (Chair); Professor Bertram Ludäscher; Associate Professor Vetle I. Torvik; Daniel C. Miller, Associate Professor in the Keough School of Global Affairs at the University of Notre Dame.
Abstract
Understanding funding dynamics in Library and Information Science (LIS) is crucial for revealing how knowledge production, research collaboration, and institutional behaviors are shaped. However, inconsistencies and ambiguities in funding metadata hinder large-scale analyses. This study develops scalable NLP approaches for Organization Name Disambiguation (OND), integrating traditional machine learning and large language models (LLMs) with transfer learning and in-domain learning strategies on a human-annotated OND gold dataset. The disambiguated dataset enables comprehensive analyses of LIS funding patterns over the past decade, including descriptive statistics, co-funding network visualizations, and public-private collaboration trends using LLMs. To support cross-domain comparison, a parallel analysis is conducted in the Biodiversity Conservation (BC) domain. By combining OND and funder analyses, this research provides new insights into the LIS and BC funding landscape and contributes methodological advances applicable to other domains. As funding data becomes increasingly central to scholarly evaluation and decision-making, developing accurate and scalable methods for organization-level analysis is both timely and necessary.
Questions? Contact Jinmo Kim.