Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.17445 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915256656199680 |
|---|---|
| author | Lieb, Anna Arora, Maneesh Mustafaraj, Eni |
| author_facet | Lieb, Anna Arora, Maneesh Mustafaraj, Eni |
| contents | Unsupervised machine learning techniques, such as topic modeling and clustering, are often used to identify latent patterns in unstructured text data in fields such as political science and sociology. These methods overcome common concerns about reproducibility and costliness involved in the labor-intensive process of human qualitative analysis. However, two major limitations of topic models are their interpretability and their practicality for answering targeted, domain-specific social science research questions. In this work, we investigate opportunities for using LLM-generated text augmentation to improve the usefulness of topic modeling output. We use a political science case study to evaluate our results in a domain-specific application, and find that topic modeling using GPT-4 augmentations creates highly interpretable categories that can be used to investigate domain-specific research questions with minimal human guidance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_17445 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Creating Targeted, Interpretable Topic Models with LLM-Generated Text Augmentation Lieb, Anna Arora, Maneesh Mustafaraj, Eni Computation and Language Unsupervised machine learning techniques, such as topic modeling and clustering, are often used to identify latent patterns in unstructured text data in fields such as political science and sociology. These methods overcome common concerns about reproducibility and costliness involved in the labor-intensive process of human qualitative analysis. However, two major limitations of topic models are their interpretability and their practicality for answering targeted, domain-specific social science research questions. In this work, we investigate opportunities for using LLM-generated text augmentation to improve the usefulness of topic modeling output. We use a political science case study to evaluate our results in a domain-specific application, and find that topic modeling using GPT-4 augmentations creates highly interpretable categories that can be used to investigate domain-specific research questions with minimal human guidance. |
| title | Creating Targeted, Interpretable Topic Models with LLM-Generated Text Augmentation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2504.17445 |