Penelitian ini bertujuan menganalisis sentimen pengguna terhadap aplikasi GoPartner berdasarkan ulasan di Google Play Store. Aplikasi GoPartner merupakan platform yang digunakan oleh mitra pengemudi Gojek untuk mengelola aktivitas harian mereka. Masalahnya adalah tingginya tingkat keluhan mitra driver pada aplikasi GoPartner yang berpotensi menurunkan kualitas pelayanan pengantaran. Permasalahan tersebut antara lain adalah alamat pelanggan tidak muncul pada aplikasi driver, saldo Gopay yang mendapatkan tips dari customer tidak langsung masuk pada akun mitra driver gojek yaitu GoPartner. Data di dapatkan melalui scraping kemudian diproses melalui tahap preprocessing teks, termasuk case folding, slang word, tokenizing, filtering, dan stemming., dengan jumlah data awal 2000. Hasil Penelitian tersebut dapat menunjukkan bahwa metode SVM mampu mengklasifikasikan ulasan pengguna dengan Tingkat akurasi yang tinggi. Dari analisis sentimen aplikasi ini ditemukan banyak ulasan pengguna bersifat positif ,namun terdapat beberapa aspek yang masih membutuhkan perbaikan. Setelah dilakukan pembersihan data, tersisa 1198 data yang digunakan dalam penelitian. Data tersebut terdiri dari 500 data positif, 364 data negatif, dan 334 data netral. Tahap pra-pemrosesan mencakup pembersihan data teks, seperti penghapusan tanda baca, angka, dan kata-kata yang tidak relevan, serta normalisasi teks. Hasil sentimen dengan metode SVM menunjukkan akurasi sebesar 79%, sebagaimana ditunjukkan oleh confusion matrix.

This research aims to analyze user sentiment towards the GoPartner application based on reviews on the Google Play Store. The GoPartner application is a platform used by Gojek driver partners to manage their daily activities. The problem is the high level of complaints from driver partners on the GoPartner application which has the potential to reduce the quality of delivery services. These problems include the customer's address not appearing on the driver application, the Gopay balance that gets tips from the customer does not go directly to the Gojek driver partner account, namely GoPartner. Data was obtained through scraping and then processed through the text preprocessing stage, including case folding, slang words, tokenizing, filtering, and stemming, with an initial amount of data of 2000. The results of this research can show that the SVM method is able to classify user reviews with a high level of accuracy. . From the sentiment analysis of this application, it was found that many user reviews were positive, but there were several aspects that still needed improvement. After data cleaning, 1198 data remained for use in the research. The data consists of 500 positive data, 364 negative data and 334 neutral data. The pre-processing stage includes cleaning text data, such as removing punctuation, numbers and irrelevant words, as well as text normalization. The sentiment results using the SVM method show an accuracy of 79%, as shown by the confusion matrix.