میزان همخوانی کلیدواژه‌های مستخرج از چکیده با توصیفگرهای نمایه‌سازان در پایگاه «چکیده پایان‌نامه‌های ایران»

Purpose: This research is devoted to study the consistency between keywords extracted from abstracts of theses by the experts in the related fields and descriptors provided by the indexers in database of “Iran’s theses abstracts”. Methodology: This research is an applied study based on content analysis. A checklist which consisted of 32 criteria was used. In addition, we consulted the experts in the studied fields. The population under study included the records in database of “Iran’s theses abstracts” (74,500 records) during 1989-2006. By using the sample size formula, 100 records were chosen from each of the fields of human sciences, basic sciences, engineering, and agriculture, by random stratified sampling. Findings: According to results of the research, there are significant relationships between the number of keywords in theses abstracts and their consistency with the descriptors, as well as between the number of keywords in theses abstracts and the number of descriptors. Also, there is significant relationship in different fields and different time intervals in consistency between the extracted keywords and the descriptors. Higher rate of retrieval through descriptors in comparison with the keywords extracted from abstracts revealed the efficiency of works which have been performed in this database in order to increase the relationships between descriptors of the same subject.  Originality/Value: The findings of this research can assist the indexers and managers of the database of “Iran’s theses abstracts” for identification of strengths and weaknesses in indexing the theses and collaboration with fields experts in documents indexing.