Paper Recommendation for Research References in Data Mining using Content-Based Filtering

The recommendation system predicts items that can identify and provide high-potential content selected by the user based on the filtering information. In this study, a reference recommendation system was made to recommend reading papers for students in their final project. This system makes it easier for students to find the correct absolute assignment reading reference according to the topic taken. This study uses the content-based filtering method and the TF -IDF algorithm to find the availability of existing content. Previous studies have researched reading recommendations for students working on their final project using the Collaborative filtering method. The method requires a rating parameter to recommend something so that if there is a new item, the item is not recommended. However, there is a possibility that the thing from each content is similar. Therefore, this research will answer this problem by creating a recommendation system using Content-Based Filtering. After implementation, we tested the accuracy by dividing the test data and training data differently. System testing is done by using a confusion matrix. The results that have been tested get an accuracy of 78%. Further research suggests adding paper data and themes in recommending reading references to make the recommendations more optimal.