Agentopic: A Generative AI Agent Workflow for Explainable Topic Modeling

arXiv cs.LG / 5/5/2026

📰 NewsSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

Agentopic is a new agent-based workflow for explainable topic modeling that uses LLM reasoning to make topic assignments more transparent than approaches like LDA and BERTopic.
The method coordinates multiple agents to handle topic identification, validation, hierarchical grouping, and natural-language explanations, enabling users to trace why topics were assigned.
In experiments seeded with BBC dataset topics, Agentopic reportedly achieves an F1-score of 0.95, comparable to GPT-4.1 and better than LDA (0.93), while remaining close to BERTopic (0.98).
Agentopic can also generate large numbers of semantically coherent topics when unseeded, producing 2,045 topics across six hierarchical levels and augmenting dataset richness with generated explanations.
By embedding interpretability throughout the workflow, Agentopic is positioned as a more explainable alternative to black-box models for high-stakes domains such as finance and healthcare.

Abstract

Agentopic is a novel agent-based workflow for explainable topic modeling that leverages the reasoning capabilities of Large Language Models (LLMs). Existing topic modeling approaches such as Latent Dirichlet Allocation (LDA) and BERTopic often lack transparency on how topics are assigned or grouped. Agentopic addresses this by using multiple agents that collaboratively perform topic identification, validation, hierarchical grouping, and natural language explanation. This design enables users to trace the reasoning behind topic assignments, enhancing interpretability without sacrificing accuracy. When seeded with topics from the British Broadcasting Corporation (BBC) dataset, Agentopic achieves an F1-score of 0.95, matching GPT-4.1, improving on LDA (0.93), and close to BERTopic (0.98). We used Agentopic to augment the BBC dataset with generated explanations to improve the dataset's richness and context. The unseeded Agentopic generated 2045 semantically coherent topics organized across six hierarchical levels, vastly enriching the original five-category structure. By embedding explainability throughout the workflow, Agentopic offers an interpretable alternative to black-box models, making it particularly valuable for crucial applications like finance and healthcare.