ملخص الدراسة:

The growing requirement on the Internet have made users access to the information expressed in a language other than their own , which led to Cross lingual information retrieval (CLIR) .CLIR is established as a major topic in Information Retrieval (IR). One approach to CLIR uses different methods of translation to translate queries to documents and indexes in other languages. As queries submitted to search engines suffer lack of untranslatable query keys (i.e., words that the dictionary is missing) and translation ambiguity, which means difficulty in choosing between alternatives of translation. Our approach in this thesis is to build and develop the software tool (MORTAJA-IR-TOOL) , a new tool for retrieving information using programming JAVA language with JDK 1.6. This tool has many features, which is develop multiple systematic languages system to be use as a basis for translation when using CLIR, as well as the process of stemming the words entered in the query process as a stage preceding the translation process. The evaluation of the proposed methodology translator of the query comparing it with the basic translation that uses readable dictionary automatically the percentage of improvement is 8.96%. The evaluation of the impact of the process of stemming the words entered in the query on the quality of the output process in the retrieval of matched data in other process the rate of improvement is 4.14%. Finally the rated output of the merger between the use of stemming methodology proposed and translation process (MORTAJA-IR-TOOL) which concluded that the proportion of advanced in the process of improvement in data rate of retrieval is 15.86%. Keywords: Cross lingual information retrieval, CLIR, Information Retrieval, IR, Translation, stemming.

توثيق المرجعي (APA)