IMPLEMENTASI PENCARIAN FULL TEXT PADA DATA BERITA ONLINE MENGGUNAKAN SISTEM INDEXING APACHE SOLR; IMPLEMENTATION OF FULL TEXT SEARCH ON ONLINE NEWS DATA USING INDEXING SYSTEM OF APACHE SOLR
RUSFANDI, HENDI, I Gede Mujiyatna
2016 | Skripsi | FMIPAMedia Monitoring is an activity to monitor news published by online media. Monitoring activity is usually done by companies to know publics’ reaction over offered goods or service. To assist the monitoring activity, an application is built to help collecting, searching, reviewing, and reporting the news. Big online news data makes some processes of the application disrupted. One of the problems that occur is news searching process on the application that is slowing down. This problem occurs because of high read and write transaction on the application’s database. To reduce the database workload, the news searching process should be done outside of the database. Techniques that can be used are to do news indexing outside the database and to search the news using full text search technique. In this research, Apache Solr is used as a search server. Performance analysis of full text search between MySQL and Solr by testing with various full text keyword. An analysis of delta indexing and real-time indexing method is also conducted by simulating the addition of news data to database, periodically update the index and doing full text searching to meet Media Monitoring application’s needs. The news search result produced a result that the full text searching on Solr is 2,5 times faster than MySQL. The real-time indexing method is more suitable to implement on the Media Monitoring application that has rapid data growth, requires up-to-date data availability, and optimal use of hardware resource.
Kata Kunci : indexing, full text search, news search, MySQL indexing, Solr, realtime indexing, Media Monitoring
