Application of data mining methods on university library data

Today, most of the data generated in each field is stored in databases or data warehouses as a data stack in unstructured form. This makes it difficult to extract information from large amounts of data. Management of large unstructured amounts of user-related data in academic libraries has also become difficult and important. In this context, it has become important to conduct studies on students ' book-borrow behavior in the university library using data mining methods. In this study, data mining methods were used in the data set taken from the University Library by using information such as gender, age, department, class and which books were taken and how long the books were kept in them. First of all, the steps of cleaning up the raw data, removing unnecessary data and transferring the useful data to the analysis were provided. It is aimed to make the data meaningful and relatable by organizing raw data in irregular and complex states and creating homogeneous sets. Classification and clustering algorithms of data mining were applied during the model selection phase. Later, with the help of the weka program, the results were visualized using the students ' department, faculty/Institute, number of return days and the type of book received as criteria. At the same time, the data was studied using algorithms that best fit the purpose of the study from classification and clustering algorithms. At the end of this research, students ' book borrow behavior was examined, which books they preferred according to their departments, which courses were more researched, comparison of the books received with each other and the distribution of the books according to departments and faculties throughout the university, such as statistical results were shown with graphics. Book proposal systems that can be developed with the information and patterns obtained at the end of this study may become one of the solutions for university libraries with large volumes of books with intensive users. Using a book recommendation system in a library based on a user profile borrowing system that can also be developed will facilitate book access.