Automated Alignment and Extraction of a Bilingual Ontology for Cross-Language Domain-Specific Applications

Computational Linguistics and Chinese Language Processing Vol. 10, No. 1, March 2005, pp. 35-52 35 © The Association for Computational Linguistics and Chinese Language Processing Automated Alignment and Extraction of a Bilingual Ontology for Cross-Language Domain-Specific Applications Jui-Feng Yeh∗, Chung-Hsien Wu∗, Ming-Jun Chen∗ and Liang-Chih Yu∗ Abstract This paper presents a novel approach to ontology alignment and domain ontology extraction from two existing knowledge bases: WordNet and HowNet. These two knowledge bases are automatically aligned to construct a bilingual ontology based on the co-occurrence of words in a bilingual parallel corpus. The bilingual ontology achieves greater structural and semantic information coverage from these two complementary knowledge bases. For domain-specific applications, domain-specific ontology is further extracted from the bilingual ontology using the a island-driven algorithm and domain-specific corpus. Finally, domain-dependent terminology and axioms between domain terminology defined in a medical encyclopedia are integrated into the domain-specific ontology. In addition, a metric based on a similarity measure for ontology evaluation is also proposed. For evaluation purposes, experiments were conducted comparing an automatically constructed ontology with a benchmark ontology constructed by ontology engineers or experts. The experimental results show that the constructed bilingual domain-specific ontology mostly coincided with the benchmark ontology. As for application of this approach to the medical domain, the experimental results show that the proposed approach outperformed the synonym expansion approach to web search. Keywords: Ontology, island driven algorithm, cross language application, WordNet, HowNet ∗ Department of Computer Science and Information Engineering, National Cheng Kung University, Tainan, Taiwan, ROC E-mail: {jfyeh, chwu, mjchen, lcyu}@csie.ncku.edu.tw 36 Jui-Feng Yeh et al. 1. INTRODUCTION In the past few decades, a considerable number of studies been invested focused on developing concept bases for building technology that allows knowledge reuse and sharing. As information exchangeability and communication becomes increasingly global, multilingual lexical resources that provide transnational services are becoming increasingly important. On the other hand, multi-lingual ontologies are very important for natural language processing, such as machine translation (MT), web mining [Oyama et al. 2004], and cross-language information retrieval (CLIR). Generally, a multi-lingual ontology maps the keywords of one language to another language, or computes the co-occurrence of the words among languages. A key merit of a multilingual ontology is that it can achieve greater relation and structural information coverage by aligning or merging two or more language-dependent ontologies with different semantic features. In recent years, significant effort has focused on constructing ontologies manually according to domain experts’ knowledge. Manual ontology merging using conventional editing tools without intelligent support is difficult, labor intensive, and error prone. Therefore, several systems and frameworks to help knowledge engineers perform ontology merging have recently been proposed [Noy and Musen 2000]. To avoid reiteration in ontology construction, algorithms for ontology merging [UMLS http://umlsks.nlm.nih.gov] [Langkilde and Knight 1998] and ontology alignment [Vossen and Peters 1997] [Weigard and Hoppenbrouwers 1998] [Asanoma 2001] have been investigated. In these approaches, the final ontology is a merged version of the original ontologies with aligned links between them [Daudé et al. 2003]. Alignment is usually performed …

Content truncated to 500 words