Search Engine for Sragen District Archives System

Abstract

Sragen district archives system is the system that used to store archived data. In this archives system, the searching is based on data or a summary disposition from the archives. Because this search result is limited, this research makes search engine for regional archives system that is based on the contents of digital archives, especially Microsoft Word. This search engine is using vector space model for weighting and cosine similarity. Stemmer used in this research is stemmer based on KBBI. Evaluation is using recall and precision of queries that has been determined. The calculation of average precision (AVP) is done because of some queries are used in this evaluation. AVP measuring is calculatd based on 11 standard recall levels. Evaluation result from the search engine is relevant documents that appropriate with the sequence relevance. The value of average precision got from 10 queries is 0.896, scale from 0 to 1. The precision of the search engine performance is affected by the stemmer. Stemmer KBBI creates smaller index size, but it decreases the precision. Stemmer KBBI decreases the precision, because of search by using basic words .