<![CDATA[Fiction books are one of the most popular books in Indonesia. There are five most popular genres in fiction books, namely fantasy, mystery, romance, sci-fi, and thriller. Each genre gives a different impression and its own fans for the reader. It is common practice when people choose fiction books based on the title, author, or publisher of the book. However, this does not provide precise search results. In this final project, an application system was developed to find out fiction books based on the semantic impressions contained on the cover of the fiction book. The impression on each book cover is obtained through a survey of fiction book lovers in Indonesia. To get the results of the proximity between the user search and the impression survey data obtained through text mining, as well as the cosine similarity algorithm to calculate the most precise proximity value to the impression expected by the user. The results of this system display fiction books that have the most precise proximity value to the impression expected by the user.]]>
Copyrights © 2019
