Identifying semantic and syntactic relations from text documents

Abstract

Semantic and syntactic relations play an important role of applications in recent years, especially on Semantic Web, Information Retrieval, Information Extraction, and Question Answering. Semantic and syntactic relations content main ideas in the sentences or paragraphs. This paper presents our proposed algorithms for identifying semantic and syntactic relations between objects and their properties in order to enrich a domain specific ontology, namely Computing Domain Ontology, which is used in Information extraction system. We combine the methodologies of Natural Language Processing with Machine Learning in these proposed algorithms in order to extract the explicit and implicit relations. We exploit these relations from distinct resources, such as WordNet, Wikipedia and text documents of ACM Digital Libraries. We also use Natural Language Processing tools, such as OpenNLP, Stanford Lexical Dependency Parser in order to analyze and parse sentences. A random sample among 245 categories of ACM Categories is used to evaluate. Results generated show that our proposed approach achieves high precision.