بررسي ميزان هم خواني كليدواژه هاي عنوان و توصيفگرهاي نمايه سازان در پايگاه اطلاعاتي چكيده پايان نامه هاي ايران

A study on compatibility between title-key words and indexerʹs descriptors in "Iranian dissertations abstract" database

This research is carried out, to determine degree the of match between assigned descriptors and title keywords in "Iranian Dissertations abstracts Database". All records in this database (65539 records) considered as research population. After deleting duplicated records the research population were divided to 6 classes: Human sciences, Pure sciences, Medical sciences, Engineering sciences, Agricultural sciences and Art sciences. Sampling of each class was done according to this formula: n = NZ2 pq/ (d2 (N-1) + ZZ pq). Data gathering instrument was a check list. Chi-square test and Pearsonʹs correlation coefficient test are used for testing research questions. The results of the research show that the overlap (exact Matching) between assigned descriptors and title keyword is 45%. If we consider Relative Matching it increases to 67%. The results also show that there is a great correlation between the number of title key words and descriptors. This research also showed that 48% of assigned descriptors are uniterm and 52% are compound words. 87% of assigned descriptors are singular and 13% are in plural form.