Abstract:

Various research works have shown the requirements in terms of semantic processing to achieve more accurate Natural Language Processing (NLP) applications such as Information Retrieval (IR), Question Answering (QA), Reasoning, etc. This paper describes our approach of semantic analysis of the Arabic language. We adopt for this purpose a two-step approach; we firstly build (for once) an Arabic ontology, leveraging the content of the two linguistic resources Arabic WordNet and VerbNet and we secondly combine some Arabic NLP (ANLP) tools such as Stanford syntactic parser with the built ontology to automatically extract the semantics conveyed by the Arabic text. The extracted semantic is formulated in Conceptual Graph formalism, which is a promising and already proven powerful knowledge representation formalism, in reasoning and in many NLP fields.