Two researchers from UCL’s Department of Information Studies presented their work at the ISKO UK evening lecture series on 27 May 2026, contributing to a session themed “Research Observatory: Knowledge organization in capturing emotional semantics and oral history constructs.” The ISKO UK evening lecture series is a respected forum that brings together researchers and practitioners to explore current developments in knowledge organisation across disciplines and methodologies.
Doctoral researcher Malithi Alahapperuma presented a talk titled “Traversing the boundaries of language: Using knowledge organisation to investigate the semantics of emotion across English and Sinhala.” Malithi’s PhD investigates the design and development of large language model-based conversational tools for Sinhala, a morphologically complex low-resource language, with the broader aim of enabling speakers of under-served linguistic communities to make better use of emerging language technologies. In her presentation, Malithi addressed a fundamental challenge: putting emotions into words is never easy, and the difficulty multiplies when communicating feelings across languages, where linguistic equivalence rarely guarantees conceptual equivalence. Using English and Sinhala as a case study, her work maps emotion vocabulary across both languages through synonym rings and discourse analysis, approaching the problem through three pillars, cognition, linguistics, and socio-cultural factors to develop a more grounded framework for understanding the semantics of emotion in cross-linguistic settings.
Dr Andreas Vlachidis, Associate Professor of Information Science and Deputy Head of the Department of Information Studies, presented “Ontology Development with Large Language Models for Oral History Network Construction and Hermeneutical Analysis.” His talk approached oral history as situated, intersubjective, and performative, drawing on a corpus of 25 interviews charting the history of Digital Humanities. He introduced the Hermeneutic Ontology Engineering methodology developed within the AHRC-DFG funded MeDoraH project. The talk examined how large language models can accelerate entity and relationship detection in conceptual modelling, while critically assessing their limitations and arguing for a hybrid approach that integrates human expertise at key decision points to maintain semantic precision.
Both presentations showcased the department’s strength in knowledge organisation research and its application to complex challenges in computational humanities and cross-linguistic semantics.