A new IR expert system for document ranking in Persian databases

Getting access to relevant information in information databases and the World Wide Web is deemed as an important challenge in any IR system. the aim of this paper was to investigate the use of expert systems to improve retrieval ranking in Persian search engines. The query terms input by the user were processed by the IR system and the documents retrieved were listed based on their degree of similarity or relevance to the search terms. This type of sorting is of utmost importance to any IR system, and any mismatch in the sorting will result in failure of the user to get access to his/her information needs despite the availability of the data in the database. Findings/ The quality of the first 01 documents retrieved is crucial to the assessment of the relevant IR system. This paper intends to enhance, through the application of an expert system, the quality and efficiency of document retrieval in Persian databases search engines. The expert system computes the value and relevance of each article and hence outperforms traditional systems. This intelligent system transforms human parameters to expert system instructions and rules to assess the value of articles and, in practice, outperforms non-intelligent systems.