Academic Discourse on ChatGPT in Social Sciences: A Topic Modeling and Sentiment Analysis of Research Article Abstracts

This repository contains the dataset and code used in the study titled “Academic Discourse on ChatGPT in Social Sciences: A Topic Modeling and Sentiment Analysis of Research Article Abstracts.” The study explores how social science scholars frame and evaluate ChatGPT by analyzing 1,227 SSCI-indexed abstracts using Latent Dirichlet Allocation (LDA) topic modeling and lexicon-based sentiment analysis. The data include the collected abstracts (with metadata), while the code files provide the full analytical pipeline in Python and R, covering preprocessing, topic modeling, sentiment scoring using the NRC Emotion Lexicon, and visualization scripts. This repository supports transparency, reproducibility, and reuse of the study’s computational methods and underlying materials.