Information extraction for traffic congestion in social network: Case study: Bekasi city

The growth of the use of social networks becomes the concern of researchers and many parties to get the information contained in them based on various kinds of needs, one of which is the need to build a smart city that can monitor its traffic condition. Twitter is chosen in this study because of its update intensity about traffic congestion that is higher than those of other social networks. Bekasi City is chosen for this case study because it has sufficient potential of data source from Twitter. Data processing from the social network needs information network approach such as performing information extraction and classification. The classification is performed to distinguish between the data which are related and not related to the traffic condition using SVM (Support Vector Machine). The information extraction is performed to obtain valuable information, including location, traffic condition, congestion causes, weather condition, and time of occurrence. The experiment that has been performed shows that the information extraction and classification method that is used gives a good result and better from previous study, which is 86% for classification, 81% for traffic location, 78% for congestion causes, 96% for weather condition, and 98-100% for time of occurrence extraction. Besides that, the information extraction accompanied by the reliability value of information based on the formulation result of tweet attributes using RSVT (Reliability Support Value for Tweet).