ERSOM: A Structural Ontology Matching Approach Using Automatically Learned Entity Representation

Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 2419–2429, Lisbon, Portugal, 17-21 September 2015. c⃝2015 Association for Computational Linguistics. ERSOM: A Structural Ontology Matching Approach Using Automatically Learned Entity Representation Chuncheng Xiang, Tingsong Jiang, Baobao Chang, Zhifang Sui Key Laboratory of Computational Linguistics, Ministry of Education School of Electronics Engineering and Computer Science, Peking University Collaborative Innovation Center for Language Ability, Xuzhou 221009 China {ccxiang,tingsong,chbb,szf}@pku.edu.cn Abstract As a key representation model of knowl- edge, ontology has been widely used in a lot of NLP related tasks, such as se- mantic parsing, information extraction and text mining etc. In this paper, we study the task of ontology matching, which con- centrates on finding semantically related entities between different ontologies that describe the same domain, to solve the semantic heterogeneity problem. Previ- ous works exploit different kinds of de- scriptions of an entity in ontology di- rectly and separately to find the corre- spondences without considering the high- er level correlations between the descrip- tions. Besides, the structural informa- tion of ontology haven’t been utilized ad- equately for ontology matching. We pro- pose in this paper an ontology matching approach, named ERSOM, which main- ly includes an unsupervised representation learning method based on the deep neural networks to learn the general representa- tion of the entities and an iterative sim- ilarity propagation method that takes ad- vantage of more abundant structure infor- mation of the ontology to discover more mappings. The experimental results on the datasets from Ontology Alignment Eval- uation Initiative (OAEI1) show that ER- SOM achieves a competitive performance compared to the state-of-the-art ontology matching systems. 1The OAEI is an international initiative organizing annual campaigns for evaluating ontology matching systems. All of the ontologies provided by OAEI are described in OWL-DL language, and like most of the other participates our ERSOM also manages the OWL ontology in its current version. OAEI: http://oaei.ontologymatching.org/ 1 Introductions In the recent years, it becomes evident that one of the most important directions of improvement in natural language processing (NLP) tasks, like word sense disambiguation, coreference resolu- tion, relation extraction, and other tasks related to knowledge extraction, is by exploiting seman- tics resources (Bryl et al., 2010). Nowadays, the Semantic Web made available a large amount of logically encoded information (e.g. ontologies, RDF(S)-data, linked data, etc.), which constitutes a valuable source of semantics. However, extend- ing the state-of-the-art natural language applica- tions to use these resources is not a trivial task mainly due to the heterogeneity and the ambiguity of the schemes adopted by the different resources of the Semantic Web. How to utilize these re- sources in NLP tasks comprehensively rather than choose just one of them has attracted much atten- tion in recent years. An effective solution to the ontology hetero- geneity problem is ontology matching (Euzenat et al., 2007; Shvaiko and Euzenat, 2013), whose main task is to establish semantic correspondences between entities (i.e., classes, properties or in- stances) from different ontologies. Ontology matching is usually done …

Content truncated to 500 words