TY - GEN
T1 - Event classification in surabaya on twitter with support vector machine
AU - Ajipangestu, Drajad Bima
AU - Sarno, Riyanarto
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/9/19
Y1 - 2020/9/19
N2 - Twitter is a social media that is often used by many people in the world. The information is spread and obtained through social media. For example, there is a company that is organizing a new event that many people need to know. This allows the creation of a system that supports the presentation of user information by detecting certain events from Twitter's social media data. In this study, tweet data will be retrieved using Twitter API and stored in JSON format. Furthermore, there will be a pre-processing which includes the deletion of characters, number, URL, stemming, and lower case. Furthermore, feature extraction is performed using Global Vector for Word Representation. we will classify into four classes, which are Competitions, Seminars, Festivals, and Other events. The classification is using SVM to predict the type of event. There are three experimental methods used, there is SVM C, SVM linear, and SVM Nu. SVM Nu was conducted with changes in the SVC parameters in the form of kernel and Nu to produce the best accuracy. Based on the experiments we have done, the best results are obtained with an accuracy of 85.2% by classification using the NuSVC method with an RBF kernel and nu parameter of 0.2.
AB - Twitter is a social media that is often used by many people in the world. The information is spread and obtained through social media. For example, there is a company that is organizing a new event that many people need to know. This allows the creation of a system that supports the presentation of user information by detecting certain events from Twitter's social media data. In this study, tweet data will be retrieved using Twitter API and stored in JSON format. Furthermore, there will be a pre-processing which includes the deletion of characters, number, URL, stemming, and lower case. Furthermore, feature extraction is performed using Global Vector for Word Representation. we will classify into four classes, which are Competitions, Seminars, Festivals, and Other events. The classification is using SVM to predict the type of event. There are three experimental methods used, there is SVM C, SVM linear, and SVM Nu. SVM Nu was conducted with changes in the SVC parameters in the form of kernel and Nu to produce the best accuracy. Based on the experiments we have done, the best results are obtained with an accuracy of 85.2% by classification using the NuSVC method with an RBF kernel and nu parameter of 0.2.
KW - SVM
KW - Text Classification
KW - Twitter
UR - http://www.scopus.com/inward/record.url?scp=85096860611&partnerID=8YFLogxK
U2 - 10.1109/iSemantic50169.2020.9234205
DO - 10.1109/iSemantic50169.2020.9234205
M3 - Conference contribution
AN - SCOPUS:85096860611
T3 - Proceedings - 2020 International Seminar on Application for Technology of Information and Communication: IT Challenges for Sustainability, Scalability, and Security in the Age of Digital Disruption, iSemantic 2020
SP - 482
EP - 486
BT - Proceedings - 2020 International Seminar on Application for Technology of Information and Communication
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2020 International Seminar on Application for Technology of Information and Communication, iSemantic 2020
Y2 - 19 September 2020 through 20 September 2020
ER -