ANALISIS SENTIMEN TERHADAP LAYANAN SHOPEEFOOD PADA MEDIA SOSIAL TWITTER MENGGUNAKAN METODE NAÏVE BAYES

• Main Author: Muhamad Farhan Purnomo Adjie, (Author)
• Format: Book
• Published: 2023-01-10.
• Subjects: Thesis; NonPeerReviewed
• Online Access: Link Metadata
• Notes: http://repository.upnvj.ac.id/23280/1/ABSTRAK.pdf; http://repository.upnvj.ac.id/23280/2/AWAL.pdf; http://repository.upnvj.ac.id/23280/3/BAB%201.pdf; http://repository.upnvj.ac.id/23280/4/BAB%202.pdf; http://repository.upnvj.ac.id/23280/5/BAB%203.pdf; http://repository.upnvj.ac.id/23280/6/BAB%204.pdf; http://repository.upnvj.ac.id/23280/7/BAB%205.pdf; http://repository.upnvj.ac.id/23280/8/DAFTAR%20PUSTAKA.pdf; http://repository.upnvj.ac.id/23280/9/RIWAYAT%20HIDUP.pdf; http://repository.upnvj.ac.id/23280/10/LAMPIRAN.pdf; http://repository.upnvj.ac.id/23280/11/HASIL%20PLAGIARISME.pdf; http://repository.upnvj.ac.id/23280/12/ARTIKEL%20KI.pdf
• Keywords: QA75 Electronic computers. Computer science; QA76 Computer software; Z665 Library Science. Information Science

The many choices of online food delivery services are a problem for users when they want to use these services. So, users have their own experience of each food delivery service because it has different advantages and disadvantages, one of the ready-to-eat food delivery service providers, namely shopeefood. User experience of the shopeefood service is presented in a sentiment analysis based on opinion data from Twitter social media users, the data collection process uses Twitter's API (Application Programming Interface). This research was conducted to determine the performance of the model that was created using the Naïve Bayes method for the shopeefood service to determine the accuracy value. The research was conducted by classifying opinion data from Twitter social media users into two, namely negative sentiment and positive sentiment based on manual labeling carried out by 3 annotators which later the data will be pre-processed, TF-IDF weighted, as well as the first two data divisions of 30% test data and 70% training data, the second 20% test data and 80% training data before entering the naïve Bayes classification modeling stage. The evaluation results for the naïve Bayes classification model in the first data distribution of 30% test data and 70% training data obtained an accuracy of 71.875%, a recall of 70.27%, a precision of 74.71%, and a specificity of 73.65%. Meanwhile, in the second data division, 20% of the test data and 80% of the training data obtained an accuracy of 74.04%, a recall of 71.79%, a precision of 75%, and a specificity of 76.27%.