ANALSIS SENTIMENT REVIEW APLIKASI MOBILE LEGEND MENGGUNAKAN SUPPORT VECTOR MACHINE DAN ADABOOST

Abstrak

Game online MOBA (Multiplayer Online Battle Arena) merupakan game genre Real Time Strategy yang digandrungi player saat ini. Bahkan terdapat event turnament setiap tahunnya salah satunya game mobile legend. Pertumbuhan pengguna Mobile Legends pada tahun 2016 dimainkan sebanyak 30 juta orang, sedangkan pada tahun 2018 mencapai 200 juta orang. Pada App Store jumlah pengguna yang mendownload aplikasi pada saat ini mencapai lebih dari 500 juta download sehingga Jumlah tersebut membuat Mobile Legends selalu menduduki peringkat pertama di playstore. Pada ulasan play store sendiri terdapat 31 juta ulasan masyarakat mengenai game mobile legend tersebut. Pada ulasan masyarakat mengenai game mobile legend memeliki komentar dan pendapat masyarakat yang berbeda-beda baik itu saran, support ataupun dukungan dan tak jarang pada ulasan tersebut terdapat komentar-komentar negatif mengenai game ini. Opini-opini masyarakat tersebut dapat diolah dengan mengguanakan pembelajaran learning untuk mengetahui komentar positif dan komentar negative salah satunya menggunakan Vector Machine (SVM) dan juga Adaboost. Hasil dari pemodelan algoritma yang dilakukan mendapatkan nilai akurasi yang sangat tinggi yaitu 99,31%

The MOBA (Multiplayer Online Battle Arena) online game is a Real Time Strategy genre game that players love today. There is even a tournament event every year, one of which is the legendary mobile game. The growth of Mobile Legends users in 2016 was played by 30 million people, while in 2018 it reached 200 million people. At the App Store, the number of users who have downloaded the application currently reaches more than 500 million downloads, so that this number makes Mobile Legends always ranked first in the Playstore. In the Play Store review itself, there are 31 million public reviews about this legendary mobile game. In community reviews about the legendary mobile game, there are comments and opinions from the community that are different, be it suggestions, support or support, and not infrequently in these reviews there are negative comments about this game. Community opinions can be processed using learning methods to find out positive comments and negative comments, one of which is using Vector Machine (SVM) and also Adaboost. The results of the algorithm modeling carried out get a very high accuracy value of 99.31%