Probabilistic Topic Models for Enriching Ontology from Texts

Abstract

The ontology enrichment process is text-based and the application domain in hand is circumscribed to the content of the related texts. However, the main challenge in ontology enrichment is its learning, since there is still a lack of relevant approach able to achieve automatic enrichment from a textual corpus or dataset of various topics. In this paper, we describe a new approach for automatic learning of terminological ontologies from textual corpus based on probabilistic models. In our approach, two topic modeling algorithms are explored, namely LDA and pLSA for learning topic ontology. The objective is to capture semantic relationships between word-topic and topic-document in terms of probability distributions to build a topic ontology and ontology graph with minimum human intervention. Experimental analysis on building a topic ontology and retrieving corresponding topic ontology for a user query demonstrates the effectiveness of the proposed approach.