<![CDATA[The novel is a story that has a long, imaginary plot. Based on the editor's choice on the Amazon.com website, 50 of the 100 best-selling books are novels. This shows that public interest in the novel is quite high as one type of reading. But when you want to choose a novel that you want to read, readers sometimes feel confused to know the quality of the novel. One reference in looking at the quality of a product is rating. The Goodreads site is one site that allows amateur reviewers to write reviews and ratings to help readers choose relevant books. But sometimes Goodreads users don't give ratings to a book so followers from that user want to know the rating given by the user in the book. This study uses the Genre Based Collaborative Filtering method as a calculation of rating predictions and Text Similarity to determine the value of similarity between documents with each other. The data used in this study were 31 users and 90 synopsis as training data and 35 synopsis as test data. System accuracy obtained from the classification results by using the similarity value on text similarity of 45,714286% and MAE value of 0,27742857 so that it can be concluded that the method of genre based collaborative filtering and text similarity can be used to make rating predictions.]]>
Copyrights © 2019
