<![CDATA[A recommendation system is a software technique that provides suggestions orrecommendations based on user preferences. This paper examines arecommendation system for laptops at Els Computer Shop in Semarang, usinga content-based filtering approach to help customers select laptops that matchthe attributes they are looking for. The prototype method was chosen as thesystem development technique because it facilitates interaction between systemdevelopers and can address issues between users and analysts. The design of thisrecommendation system involves two types of data encoding: one-hot encodingand ordinal encoding. One-hot encoding transforms categorical laptop data intobinary numbers (0 or 1), while ordinal encoding converts categorical laptopdata based on a specific order into numeric values (0, 1, 2, 3). The transformeddata is then calculated using cosine similarity to determine the similarity scoreof recommended laptops. The results of the laptop recommendation systemdisplay three laptops that are most similar to the searched laptop index basedon cosine similarity calculations.]]>
Copyrights © 2024
