IDENTIFIKASI KALIMAT KUTIPAN DARI TEKS BERITA ONLINE BERBAHASA INDONESIA DENGAN METODE BERBASIS ATURAN; QUOTATIONS IDENTIFICATION FROM INDONESIAN ONLINE NEWS USING RULE-BASED METHOD

SYAIFUDIN, YUSUF, Arif Nurwidyantoro

2016 | Skripsi | FMIPA

News is a source of information that is often used as data in information extraction. Interviewees statement or opinion is one of the sources of information in the news, which is quite important to know things that are said by a person or organization related to a specific event. Interviewees statement is important because it is one of the information that can affect the mind of a newsreader. Moreover, the statement may also indicate the speaker’s sentiment related to an event that was reported. The statement is often written using quotation, by writing rules of either direct quotes or indirect quotes. This study uses a rule-based method which quotes in the written news are extracted based on common rules contained in the news. Each pattern of writing quotations found in a collection of news data is made into a class of rule that represents the pattern. Features used in the class of rules is the existence of reporting verb and entities in a sentence. The result of this study indicate that the entire class rules have been made are able to extract sentence statement pretty well. The system is able to detect sentence statement with an accuracy of 88.618%, precision of 99.013% and recall 79.936%.

Kata Kunci : quotations extraction, opinion extraction, rule-based, information extraction.