文章基本信息

标题：MUSIC INTEREST CLASSIFICATION OF TWITTER USERS USING SUPPORT VECTOR MACHINE
本地全文：下载
作者：YUSRA ; MUHAMMAD FIKRY ; BAMBANG RIYANTO TRILAKSONO 等
期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：11
出版社：Journal of Theoretical and Applied
摘要：Interest determination in music is very beneficial for business communities such as social network advertiser, music studio rental, musical instrument sales, and music concert promoter. This research discusses the classification of music interest of Twitter users based on their tweets in Bahasa Indonesia (Indonesian language). We classify tweets into three music genre categories (jazz, pop, or rock) and three sentiments (positive, negative, or neutral) using Support Vector Machine (SVM). Tweet text classification includes user text, retweet text, mention, hashtag, emoticon, and link (URL). Preprocessing is initiated with word segmentation, removal of symbol and numeric character codes, stemming, word normalization, removal of stopwords, and searching DBpedia for some important words that does not have basic words. This research use dataset of 450 tweets. By generating SVM model on training process that use 360 tweets, Gaussian RBF kernel, and 10-fold cross validation, a pair of parameter (C=0.7, γ=0.9) for music genre category and a pair of parameter (C=0.7, γ=0.8) for sentiment were obtained. The testing process use 90 tweets and resulting the best accuracy for music genre category (96.67%) and the best accuracy for sentiment (86.67%).
关键词：Music Genre Category; Sentiment; SVM; Tweet