{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/125694"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/125694","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Code generation for few-shot event structure prediction","abstract":"Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2026-08-01","abstract_html":"Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2026-08-01","abstract_has_math":false,"creators":["Wang, Xingyao"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"M.S.","degree_level":"Thesis","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Ji, Heng"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2024,"date_issued":"2024-07-08","date_published":"2024-07-08","updated_at":"2026-07-22T22:25:02Z","subjects":["Event Argument Extraction","Code Generation","Large Language Model"],"languages":["en","eng"],"rights":["Copyright 2024 Xingyao Wang"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/2142/125694","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Ji, Heng"]},{"key":"dc:creator","label":"Author","values":["Wang, Xingyao"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2024-07-08","2024-08"]},{"key":"dc:type","label":"Dc Type","values":["text","Thesis"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Event Argument Extraction","Code Generation","Large Language Model"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en","eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2024 Xingyao Wang"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://hdl.handle.net/2142/125694"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2026-08-01","The student, Xingyao Wang, accepted the attached license on 2024-07-05 at 10:18.","The student, Xingyao Wang, submitted this Thesis for approval on 2024-07-05 at 10:37.","This Thesis was approved for publication on 2024-07-08 at 16:56.","DSpace SAF Submission Ingestion Package generated from Vireo submission #20952 on 2025-02-04 at 21:16:22","Large Language Model (LLM) trained on a mixture of text and code has demonstrated impressive capability in translating natural language (NL) into structured code. We observe that semantic structures can be conveniently translated into code and propose Code4Struct to leverage such text-to-structure translation capability to tackle structured prediction tasks. As a case study, we formulate Event Argument Extraction (EAE) as converting text into event-argument structures that can be represented as a class object using code. This alignment between structures and code enables us to take advantage of Programming Language (PL) features such as inheritance and type annotation to introduce external knowledge or add constraints. We show that, with sufficient in-context examples, formulating EAE as a code generation problem is advantageous over using variants of text-based prompts. Despite only using 20 training event instances for each event type, Code4Struct is comparable to supervised models trained on 4,202 instances and outperforms current state-of-the-art (SOTA) trained on 20-shot data by 29.5% absolute F1. By leveraging the inheritance feature of PL, Code4Struct can use 10-shot training data from a sibling event type to predict arguments for zero-resource event types and outperforms the zero-shot baseline by 12% absolute F1. All code and resources are publicly available at https://github.com/xingyaoww/code4struct."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Code generation for few-shot event structure prediction"]}]}],"canonical_facts":{"dc:contributor":["Ji, Heng"],"dc:creator":["Wang, Xingyao"],"dc:date":["2024-07-08","2024-08"],"dc:description":["Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2026-08-01","The student, Xingyao Wang, accepted the attached license on 2024-07-05 at 10:18.","The student, Xingyao Wang, submitted this Thesis for approval on 2024-07-05 at 10:37.","This Thesis was approved for publication on 2024-07-08 at 16:56.","DSpace SAF Submission Ingestion Package generated from Vireo submission #20952 on 2025-02-04 at 21:16:22","Large Language Model (LLM) trained on a mixture of text and code has demonstrated impressive capability in translating natural language (NL) into structured code. We observe that semantic structures can be conveniently translated into code and propose Code4Struct to leverage such text-to-structure translation capability to tackle structured prediction tasks. As a case study, we formulate Event Argument Extraction (EAE) as converting text into event-argument structures that can be represented as a class object using code. This alignment between structures and code enables us to take advantage of Programming Language (PL) features such as inheritance and type annotation to introduce external knowledge or add constraints. We show that, with sufficient in-context examples, formulating EAE as a code generation problem is advantageous over using variants of text-based prompts. Despite only using 20 training event instances for each event type, Code4Struct is comparable to supervised models trained on 4,202 instances and outperforms current state-of-the-art (SOTA) trained on 20-shot data by 29.5% absolute F1. By leveraging the inheritance feature of PL, Code4Struct can use 10-shot training data from a sibling event type to predict arguments for zero-resource event types and outperforms the zero-shot baseline by 12% absolute F1. All code and resources are publicly available at https://github.com/xingyaoww/code4struct."],"dc:format":["application/pdf"],"dc:identifier":["https://hdl.handle.net/2142/125694"],"dc:language":["en","eng"],"dc:rights":["Copyright 2024 Xingyao Wang"],"dc:subject":["Event Argument Extraction","Code Generation","Large Language Model"],"dc:title":["Code generation for few-shot event structure prediction"],"dc:type":["text","Thesis"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["M.S."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:25:02Z"}