Main Article Content
Abstract
Information Retrieval System is a process of searching for a set of documents relevant with needs of the user. In the teaching and learning activities, Information Retrieval System much needed. For example at library, teachers and students can search for books that can be used as reference material. With the Information Retrieval System can improve search results more relevant book to the needs of the user.Library of STMIK PPKIA Tarakanita Rahmawati domiciled in Tarakan, until now have 4215 ± titles consisting of printed books and research. In this research, using 500 data books and software used to create an IR system that is Sphinx Search. The role of Sphinx Search in a IR system is the process of indexing for data set books and search for data relevant books matching the query entered by user.Methods classification text used in this research are Weight Adjusted K-Nearest Neighbor (WAK-NN). Percentage accuracy of match type classes generated by WAK-NN is 81% on 150 data test with processing time by ± 11 minutes / book. By making use of classification text, it can improve the performance of the effectiveness of the results data books are displayed by Sphinx Search. With produce value of precision 63.6% for 5 different queries. Other than displaying the data books that match the query, in this research can also display a list of books that may be used as reference material / other teachings. The establishment of documents cluster made with method Clustering Minimum Spanning Tree (MST) using the formula Cosine Distance Measure.