<![CDATA[The use of e-books on the development of digital technology makes it easy for users to get more practical information than having to use printed books. The number of e-books spread on the internet is very numerous and varied, therefore a system for retrieving information on e-book documents in Indonesian is needed. To improve the relevance of the results of the returned documents, relevance feedback techniques can be applied. Relevance feedback is a technique where users can provide feedback on previous document search results. Sorting the number of documents returned based on queries is calculated using the BM25 method. This study aims to determine the results of the application and the results of testing of relevance feedback on the retrieval system of Indonesian e-book document information using the BM25 method. Based on the test results, the AVP value after relevance feedback has decreased. In testing based on K values, the AVP value before RF is 0.592, after RF(20) is 0,558, and after RF(50) is 0.573. While in testing based on the expansion terms, the AVP value before RF is 0.593, after RF(20) is 0,587 and after RF(50) is 0.570.]]>
Copyrights © 2019
