Unsupervised Terminological Ontology Learning based on Hierarchical Topic Modeling

TL;DR

This paper introduces hrLDA, a hierarchical topic model that extracts terminological ontologies from diverse documents by leveraging noun phrases, syntax, and document structure, outperforming traditional models especially in noisy data scenarios.

Contribution

The paper presents hrLDA, a novel hierarchical relation-based LDA model that improves ontology learning by incorporating syntax and structure, and demonstrates its effectiveness over existing models.

Findings

hrLDA outperforms existing topic models in hierarchy building.

hrLDA is robust to noisy datasets.

Ontologies from hrLDA are comparable to expert-created ontologies.

Abstract

In this paper, we present hierarchical relationbased latent Dirichlet allocation (hrLDA), a data-driven hierarchical topic model for extracting terminological ontologies from a large number of heterogeneous documents. In contrast to traditional topic models, hrLDA relies on noun phrases instead of unigrams, considers syntax and document structures, and enriches topic hierarchies with topic relations. Through a series of experiments, we demonstrate the superiority of hrLDA over existing topic models, especially for building hierarchies. Furthermore, we illustrate the robustness of hrLDA in the settings of noisy data sets, which are likely to occur in many practical scenarios. Our ontology evaluation results show that ontologies extracted from hrLDA are very competitive with the ontologies created by domain experts.