Date of Award
12-2025
Culminating Project Type
Thesis
Styleguide
ieee
Degree Name
Computer Science: M.S.
Department
Computer Science and Information Technology
College
School of Science and Engineering
First Advisor
Jalal Khalil
Second Advisor
Andrew Anda
Third Advisor
Lynn MacDonald
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Keywords and Subject Headings
Artificial Intelligence, Large Language Models, Natural Language Inference, Economics
Abstract
Models of causality are cornerstones of contemporary economics. The identification and classification of economic causality performed by current artificial intelligence (AI) models applied to documents which include economic content can be inaccurate and imprecise. EconNLI is the primary document dataset serving as a corpus used for assessment of economic causality in documents with economic content. We augment EconNLI by adding documents which serve to extend and broaden the set of identifiable economic causality categories. We then assess the identification and classification of economic causality of select language models of our expanded EconNLI dataset. Our assessment metrics include precision, recall, and accuracy. We show that our dataset provides a more extended and broadened assessment tool.
Recommended Citation
Shaikh, Nazimuddin, "Extending and Broadening Economics Inference Benchmarking Dataset for Large Language Models" (2025). Culminating Projects in Computer Science and Information Technology. 65.
https://repository.stcloudstate.edu/csit_etds/65
