Large organizations contain huge structured and unstructured data. This data need to be analyzed and retrieved as a part of their daily business. Data extractor that depends on entity recognition to extract data from documents and converts it into structured database can solve the problem of searching in unstructured data. In addition, semantic search engines that use query expansion to extract results that are more informative can solve the problem of polysemy and synonymy. This paper proposes a complete solution to solve these problems. An Arabic semantic search engine is proposed which consists of four components (data extractor, taxonomy builder, database indexer, and search engine). The system is applied on a real case study of a large governmental organization's database. The results show superior performance compared to other solutions. It gives good measures for the F-score |
