MALTopic: Multi-Agent LLM Topic Modeling Framework

TL;DR

MALTopic introduces a multi-agent framework utilizing large language models to improve the extraction of coherent, diverse, and interpretable topics from survey data by integrating structured responses and specialized tasks.

Contribution

It presents a novel multi-agent LLM-based approach that enhances traditional topic modeling by incorporating structured data and task specialization for better interpretability.

Findings

Significantly improves topic coherence and diversity.

Produces more human-readable and contextually relevant topics.

Outperforms LDA and BERTopic on survey datasets.

Abstract

Topic modeling is a crucial technique for extracting latent themes from unstructured text data, particularly valuable in analyzing survey responses. However, traditional methods often only consider free-text responses and do not natively incorporate structured or categorical survey responses for topic modeling. And they produce abstract topics, requiring extensive human interpretation. To address these limitations, we propose the Multi-Agent LLM Topic Modeling Framework (MALTopic). This framework decomposes topic modeling into specialized tasks executed by individual LLM agents: an enrichment agent leverages structured data to enhance textual responses, a topic modeling agent extracts latent themes, and a deduplication agent refines the results. Comparative analysis on a survey dataset demonstrates that MALTopic significantly improves topic coherence, diversity, and interpretability compared to LDA and BERTopic. By integrating structured data and employing a multi-agent approach, MALTopic generates human-readable topics with enhanced contextual relevance, offering a more effective solution for analyzing complex survey data.