SISTEM PEREKOMENDASI DENGAN SINGULAR VALUE DECOMPOSITION DAN TEKNIK SIMILARITAS PEARSON CORRELATION

作者：Universitas Methodist Indonesia, Rimbun Siringoringo, Jamaluddin Jamaluddin, Gortab Lumbantoruan · 发表于：METHODIKA Jurnal Teknik Informatika dan Sistem Informasi · 年份：2021 · DOI：10.46880/mtk.v7i1.257 · 被引用次数：1 · 研究领域：Data Mining and Machine Learning Applications、Edcuational Technology Systems、Information Retrieval and Data Mining

The growth of e-commerce has resulted in massive product information and huge volumes of data. This results in data overload problems. In the case of e-commerce, consumers or users spend a lot of time choosing the goods they need. The urgent question to be answered at this time is how to provide solutions related to intelligent information restrictions so that the existing information is truly information that is by preferences and needs. This research performs information filtering by applying the singular value decomposition method and the Pearson similarity technique to the book recommendation system. The data used is the Book-Crossing Dataset which is the reference dataset for many research recommendation systems. The resulting recommendations are then compared with e-commerce recommendations such as amazom.com. Based on the results of the study obtained data that the results of the recommendations in this study are very good and accurate.