نوع مقاله : مقاله پژوهشی
نویسندگان
گروه مهندسی کامپیوتر و فناوری اطلاعات، دانشکده فنی و مهندسی، دانشگاه پیام نور، تهران، ایران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
With the rapid growth of digital resources and the increasing volume of information in digital libraries, users face significant challenges in locating relevant materials quickly, accurately, and in a personalized manner. Traditional search and filtering approaches, which primarily rely on keyword matching and limited user preference analysis, often fail to fully satisfy users’ information needs. Consequently, intelligent recommender systems have gained considerable attention as an effective solution for improving access to resources and enhancing user experience. This study proposes a hybrid recommender system based on behavioral and content-based data, consisting of data preprocessing, two recommendation engines (content-based and collaborative filtering), and an intelligent aggregation layer for combining recommendation outputs. Behavioral data were extracted from the Book-Crossing dataset, while content-based information was obtained from both Book-Crossing and CiteULike datasets to model user preferences and resource characteristics. The performance of the proposed model was evaluated using precision, recall, coverage, diversity, and an intelligent recommendation quality score. Experimental results demonstrated that the hybrid model outperformed single-strategy recommenders, achieving a precision of 0.75, recall of 0.72, and coverage of 0.68. Furthermore, the proposed approach reduced common recommender system challenges such as the cold-start problem and popularity bias while increasing recommendation diversity and improving user experience. The findings indicate that integrating data mining techniques with semantic content analysis provides an effective framework for developing advanced recommender systems in digital library environments.
کلیدواژهها [English]
)