Year of Graduation
Book Recommendations Based on Reading Sequences and Contextual Information
Applied Mathematics and Information Science
In this paper, algorithms of recommendation systems are used and analyzed for predicting the suffix of sequences. Information on the users’ readings of books was taken as a sequence. Information was presented by the electronic resource Bookmate. The data about perusal includes the time of reading, as well as additional data about the book, such as the genre and author of the work. For prediction, the following algorithms will be used: SVD algorithm, Recursive Neural Networks and prediction algorithm using Compact Prediction Trees. In addition, the algorithms will be refined to take into account additional information about books. Finally, the optimal input parameters of the algorithms will be selected to identify the best quality from the point of view of the selected quality metrics on this sample.