KLASIFIKASI TWEET DAN PENGENALAN ENTITAS BERNAMA PADA TWEET BENCANA DENGAN SUPPORT VECTOR MACHINE; TWEET CLASSIFICATION AND NAMED ENTITY RECOGNITION IN DISASTER TWEET WITH SUPPORT VECTOR MACHINE

Dermawan, Rizki, Edi Suharyadi

2016 | Skripsi | FMIPA

Indonesia is a country that is often affected by disaster. When there is a disaster in an area, many users of social media, especially Twitter users, provide information related to the disaster. The information provided can be the location of a disaster, the conditions of an area, or the needs of the people in the disaster area. Such information can be used for mapping the event of disaster or mapping needs of the community in the event of a disaster. However, social media has low credibility as an information provider. Moreover, unstructured of data on social media will make data collection of location, condition, and needs of the community becomes difficult. In this research, system that can classify whether a tweet related to the disaster or not has been built. If the tweet is related to the disaster, the named entity is recognized so the location of disaster, a condition of the area, and the need of community is known. Tweet classification system and named entity recognition system built by supervised learning. Supervised learning algorithm that is used is a support vector machine. In the classification of the tweet, two methods of weighting tf and tf - idf has been tried. In this study, tf

Kata Kunci : support vector machine, named entity recognition, social media, disaster.