Construction of Topic Hierarchy with Subtree Representation for Knowledge Graphs

• Published in: Axioms Vol. 14; no. 4; p. 300
• Main Authors: Zhang, Yujia, Xu, Wenjie, Yu, Zheng, Reformat, Marek Z.
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 15.04.2025
• Subjects: Algorithms; Analysis; Cluster analysis; Clustering; Data mining; Datasets; Experiments; Graphical representations; Graphs; Knowledge bases (artificial intelligence); knowledge graphs; Knowledge management; Knowledge representation; Learning models (Stochastic processes); Methods; nested hierarchical Dirichlet process; Parameter estimation; Probabilistic models; Semantics; topic modeling; Trees (Graph theory); Canada
• ISSN: 2075-1680; 2075-1680
• DOI: 10.3390/axioms14040300

Hierarchy analysis of the knowledge graphs aims to discover the latent structure inherent in knowledge base data. Drawing inspiration from topic modeling, which identifies latent themes and content patterns in text corpora, our research seeks to adapt these analytical frameworks to the hierarchical exploration of knowledge graphs. Specifically, we adopt a non-parametric probabilistic model, the nested hierarchical Dirichlet process, to the field of knowledge graphs. This model discovers latent subject-specific distributions along paths within the tree. Consequently, the global tree can be viewed as a collection of local subtrees for each subject, allowing us to represent subtrees for each subject and reveal cross-thematic topics. We assess the efficacy of this model in analyzing the topics and word distributions that form the hierarchical structure of complex knowledge graphs. We quantitatively evaluate our model using four common datasets: Freebase, Wikidata, DBpedia, and WebRED, demonstrating that it outperforms the latest neural hierarchical clustering techniques such as TraCo, SawETM, and HyperMiner. Additionally, we provide a qualitative assessment of the induced subtree for a single subject.