2026-09-20T22:30:08Z https://www.ideals.illinois.edu/oai-pmh

oai:www.ideals.illinois.edu:2142/125722 2026-02-03 col_2142_5131 col_2142_10761 com_2142_5130 com_2142_10755 com_2142_234

application/pdf en text Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2026-08-01 The student, Linyi Ding, accepted the attached license on 2024-07-11 at 20:50. The student, Linyi Ding, submitted this Thesis for approval on 2024-07-11 at 21:01. This Thesis was approved for publication on 2024-07-15 at 14:58. DSpace SAF Submission Ingestion Package generated from Vireo submission #21081 on 2025-02-04 at 21:17:04 Topic-oriented open relation extraction with seed generation Ding, Linyi 2024-07-15 Han, Jiawei Relation Extraction Large Language Models eng The field of open relation extraction (ORE) has recently observed significant advancement thanks to the growing capability of large language models (LLMs). Nevertheless, challenges persist when ORE is performed on specific topics. Existing methods give sub-optimal results in five dimensions: factualness, topic relevance, informativeness, coverage, and uniformity. To improve topic-oriented ORE, we propose a zero-shot approach called PriORE: Open Relation Extraction with a Priori seed generation. The PriORE leverages the built-in knowledge of LLM to maintain a dynamic seed relation dictionary for the topic (which is initiated a priori). It initially generates seed relations from topic-relevant entity types and can be expanded during extraction. PriORE then converts the more random ORE task to a more robust relation classification task by comparing the relation dictionary to contexts. Experiments demonstrate this approach empowers better topic-oriented control over the generated relations and thus greatly improves ORE performance along the five dimensions, especially on specialized and narrow topics. 2024-08 Thesis https://hdl.handle.net/2142/125722 Copyright 2024 Linyi Ding Siebel Computing &DataScience Computer Science University of Illinois at Urbana-Champaign M.S. Thesis