ABSTRAK
signifikan akan transportasi udara. Maskapai Lion Group secara konsisten telah mendominasi penerbangan domestik selama tiga tahun berturut-turut dengana 60% penumpang lebih banyak dari maskapai lain. Melalui media sosial Instagram, masyarakat menyampaikan keluh kesah maupun apresiasi terhadap maskapai Lion Group. Perlu dilakukan klasifikasi terhadap komentar termasuk kedalam opini apa. Dalam penelitian ini terdapat kelas sentimen positif dan negatif. Hasil analisis sentimen dapat memberikan wawasan tentang respon publik terhadap maskapai dan diharapkan dapat berkontribusi pada pemahaman yang lebih baik tentang persepsi masyarakat serta membantu Lion Group dalam meningkatkan pengalaman pelanggan dan membangun citra positif, sehingga dilakukan analisis sentimen menggunakan algoritma Support Vector Machine terhadap data komentar Instagram pada akun resmi Batik Air (@batik air), Lion Air (@lionairgroup) dan Wings Air (@wingsairid) dengan total data sebanyak 1.539 data dan dilakukan uji cross validation menggunakan nilai k-fold = 5. Hasil penelitian ini menunjukkan rata-rata sentimen positif sebesar 46.34% dan rata-rata sentimen negatif sebesar 53.66% dengan uji performa yang menghasilkan nilai rata-rata akurasi 84.93%, rata-rata presisi 85.97% dan rata-rata recall 81.32%. Hasil pembobotan Term Frequent-Inverse Dokument Frequent menghasilkan 10 kata dengan bobot tertinggi, pada sentimen positif berisi optimisme seperti kata ’buka’, ’rute’, ’baru’, ’semoga’, ’sukses’, dan pada sentimen negatif berisi keluhan mengenai ’bagasi’, ’cancel’, ’reschedule’, ’tiket’, ’admin’ dan ’pesan’.

ABSTRACT
Indonesia as the largest archipelagic country in the world, has a significant demand for air transportation. Lion Group Airlines has consistently dominated domestic flights for three consecutive years, with 60% more passengers than other airlines. Through the social media platform Instagram, the public expresses their grievances and appreciation towards Lion Group Airlines. It is necessary to classify comments into different opinions. This research focuses on positive and negative sentiment classes. The sentiment analysis results provide insights into public response towards the airlines and contribute to a better understanding of public perceptions. This analysis aims to assist Lion Group Airlines in improving customer experience and building a positive image. Sentiment analysis is performed using the Support Vector Machine algorithm on Instagram comments from the official accounts of Batik Air (@batikair), Lion Air (@lionairgroup), and Wings Air (@wingsairid), with a total of 1,539 data samples. Cross-validation is conducted using a k-fold value of 5. The research findings show an average positive sentiment of 46.34% and an average negative sentiment of 53.66%. The performance evaluation yields an average accuracy of 84.93%, average precision of 85.97%, and average recall of 81.32%. The Term Frequency-Inverse Document Frequency (TF-IDF) weighting produces the top 10 words with the highest weights. In positive sentiment, these words include 'buka' (open), 'rute' (route), 'baru' (new), 'semoga' (hopefully), and 'sukses' (success). In negative sentiment, the words include 'bagasi' (baggage), 'cancel' (cancel), 'reschedule' (reschedule), 'tiket' (ticket), 'admin' (admin), and 'pesan' (order).