PERANCANGAN PERINGKAS BERITA OTOMATIS DENGAN MEMPERHATIKAN SINONIM MENGGUNAKAN METODE WEIGHT OF FEATURE

Abstract

This application is designed for automatic news summarization by using Weight of Feature method combined with NER, sentence position, finding similarities between title and sentences, and Synonyms. When there word are 2 sentences which have  synonymous, 2 word those sentences are considered to have a closer relationship than 2 sentences that have no synonyms. This application works with extraction system and uses single document. Summarization results with Q&A evaluation result of 25% successfully answered questions with 56.24%, while 50% successfully answered questions with 83.84%, and 75% % successfully answered questions with 96.02%