Identifying significant single phrases in submitted free - Order arabic natural language questions

Abstract

In this paper, we propose a technique for extracting significant phrases from Arabic natural language questions to databases. An Arabic natural language (ANL) question is accepted as an input and then outputs all possible relations and its corresponding attributes. Arabic morphological, ontological, and syntactical analyses were applied in this paper. A lexicon is derived from the database which consists of all forms of database artifacts through morphological and ontological analysis for each relation and attribute in the database. In addition the lexicon contains all possible synonyms for each artifact in the database. The system exhibits a high degree of autonomy since it can be applied to different domains by plugging the new data elements of the database and the system will process the database schema and generate the lexicon.