Purpose: The main purpose of this study is to design and evaluate a book recommender system in digital and public libraries. The solution has been provided by receiving and reviewing the preferences and experiences of users and profile information and studying the background of each user, as well as considering groups of features recorded in the recommendation process. Method: This research is applied in terms of purpose and survey method. The statistical population studied in this research consists of 263 questionnaires of users and 30 questionnaires of librarian experts. In order to find similarity between users and books, clustering and grouping have been used. Findings: There are two criteria for grouping: users grouping that can be used on the three indicators of age, gender, educational level, and thematic classification of books can be based on scope, branch, and sub-category. In analyzing the data in the descriptive statistics section, Excel software is used and in the analytical section, SPSS software. Findings indicate that the accuracy criterion has been improved by calculating MAE and RSME in the proposed method compared to the basic method in this field. The results also showed that classification can have a significant impact on the forecast and performance of book forecasting systems. Conclusion: The evaluation of the conceptual design showed that by focusing on user characteristics and obtaining real feedback of Iranian libraries, the recommender can serve as a key and effective element in the service of the Iranian readership community and play a good role as a virtual reference librarian.
