To access this work you must either be on the Smith College campus OR have valid Smith login credentials.

On Campus users: To access this work if you are on campus please Select the Download button.

Off Campus users: To access this work from off campus, please select the Off-Campus button and enter your Smith username and password when prompted.

Non-Smith users: You may request this item through Interlibrary Loan at your own library.

Publication Date

2025-5

First Advisor

Alicia M. Grubb

Document Type

Honors Project

Degree Name

Bachelor of Arts

Department

Computer Science

Keywords

software engineering, traceability, privacy, LLMs

Abstract

Software developers rely on the specifications contained in natural language requirements to guide their decisions when designing solutions to issues and reviewing pull requests. However, given that large projects are often specified by hundreds to thousands of requirements, the process of manually searching for relevant requirements can be very time-consuming. Large language models (LLMs) have the potential to encode the meaning of textual artifacts, such as requirements, issues, and pull requests. For small companies, supplying requirements and issues to cloud-based LLMs is not an option, as it reveals proprietary information or has significant infrastructure costs. We therefore utilize locally-run, open-source LLMs to output semantic embeddings of requirements and issues while preserving the privacy of the stakeholder. We contribute a pipeline for clustering the semantic embeddings of requirements, aiming to group requirements based on the features to which they refer, and then linking issues to clusters of requirements based on similarity. We also contribute a dataset of linked requirements and issues for evaluating the success of our pipeline in linking issues and requirements, and explore how variations on our pipeline affect its performance.

Rights

©2025 Karenna Kung. Access limited to the Smith College community and other researchers while on campus. Smith College community members also may access from off-campus using a Smith College log-in. Other off-campus researchers may request a copy through Interlibrary Loan for personal use.

Language

English

Comments

[6], 54, [5] pages: charts. Includes bibliographic references.

Download

Smith Only:
Off Campus Download

Share

COinS