<![CDATA[The use of job vacancy platforms is often used by job seekers. After users utilize the platform or job vacancy application, quite a few users leave reviews on the application so it is necessary to conduct a study regarding sentiment analysis of job vacancy applications using Lexicon based features and Support vector machines (SVM). The aim of this research is to find out the results of applying the Lexicon based method features and support vector machine for sentiment analysis of job vacancy applications. The research method used is Lexicon based features and Support vector machine with 3 research objects, namely Glints, Pintarnya and Kupu. Data collection was carried out through a scraping process on job vacancy applications in the Google Play Store. The collected data will be processed, carried out by TF-IDF, studied using Lexicon based features and support vector machines and then evaluated so that research results can be obtained that can be accounted for. The research results show that the approach through a combination of TF-IDF, Lexicon based features, and SVM with hyperparameter tuning produces good performance in sentiment analysis on text data with accuracies of 81% (Glints), 76% (Butterfly), and 82% for the Smart applicationKeywords: Sentiment Analysis; Job Vacancies Application; Lexicon Based Features; Support Vector Machine AbstrakSetelah pengguna memanfaatkan platform atau palikasi lowongan kerja dalam mencari kerja, tidak sedikit para pengguna meninggalkan ulasan pada aplikasi tersebut sehingga perlu dilakukan kajian mengenai analisis sentimen terhadap aplikasi lowongan kerja tersebut. Tujuan penelitian ini adalah untuk mengetahui hasil dari penerapan metode Lexicon based features dan Support Vector Machine (SVM) terhadap analisis sentimen aplikasi lowongan kerja. Metode penelitian yang digunakan ialah Lexicon based features dan Support vector machine dengan 3 objek penelitian yaitu Glints, Pintarnya dan Kupu. Pengumpulan data dilakukan melalui proses scraping pada aplikasi lowongan kerja yang ada di Google play store. Data yang terkumpul akan di proses, dilakukan TF-IDF, dikaji melalui Lexicon based features dan Support vector machine kemudian dievaluasi sehingga diperoleh hasil penelitian yang dapt dipertanggung jawabkan. Hasil penelitian menunjukkan bahwa pendekatan melalui kombinasi TF-IDF, fitur berbasis leksikon, dan SVM dengan penyetelan hyperparameter menghasilkan performa yang baik dalam analisis sentimen pada data teks dengan akurasi 81% (Glints), 76% (Kupu), dan 82% untuk aplikasi Pintarnya.]]>
