Research on the Impact Evaluation of Academic Monographs Based on Large Language Models

Based on the abstract text data of 2,248 sociology monographs included in the Dimensions database from 2014 to 2023, the study scores the abstracts using the ChatGPT-4o model and conducts empirical analyses by combining citation counts, Altmetrics data, and readability indices. The results show that the ChatGPT-4o scores have a strong correlation with the number of citations and Altmetrics indicators, outperforming the correlation between the readability index and citations/Altmetrics. This study provides a new framework for the textual analysis of academic monographs and demonstrates the potential of large language models in academic evaluation.