An Analysis Of Public Sentiment Toward E-Commerce On Twitter Using The Support Vector Machine (Svm) Method (A Case Study Of Shopee And Tokopedia)

Abstract

E-Commerce is electronic commerce where consumers directly buy goods from sellers through a website on the internet where transactions are carried out without intermediary services. Quoted from the iprice.co.id page (Iprice, 2022), Tokopedia and Shopee are ranked 1 and 2 in tweets on Twitter in the first quarter of 2022 with the number of tweets on Twitter reaching 1,000,000 tweets for E-Commerce Tokopedia and 778,100 tweets for Shopee. As an online store with the largest number of users and the top ranking, it must have very diverse responses, both positive, negative, and neutral from users. Therefore, sentiment analysis was carried out to determine the public's response to e-commerce shopee and tokopedia. The research data was taken from the public opinion of Twitter social media users with the hashtags #Shopeecare and #Tokopediacare then used the Support Vector Machine (SVM) algorithm as a classification method and Term Frequency - Inverse Document Frequency (TF-IDF) for word weighting in the research dataset. This study resulted in 10423 (39.6%) positive sentiments, 857 (3.25%) negative sentiments and 15059 (57.2%) neutral sentiments on the Shopee e-commerce dataset, and then 7047 (41%), 539 positive sentiments. (3.13%) and 9619 (55.9%) neutral sentiments on the Tokopedia e-commerce dataset. Accuracy results using the Support Vector Machine classification are 94% with the highest precision value of 94%, the highest recall value of 95%, and the highest f1-score value of 95% in the Shopee e-commerce dataset then produce an accuracy value of 93% with the highest precision value of 94%, the highest recall value is 94%, and the highest f1-score is 94% in the Tokopedia e-commerce dataset.