PENERAPAN PRINCIPAL COMPONENT ANALYSIS DAN GENETIC ALGORITHM PADA ANALISIS SENTIMEN REVIEW PENGIRIMAN BARANG MENGGUNAKAN ALGORITMA SUPPORT VECTOR MACHINE

Hilda Rachmi - AMIK BSI Bogor

EVOLUSI : Jurnal Sains dan Manajemen
Vol 5, No 2 (2017): Jurnal Evolusi 2017

Abstract - Sentiment analysis studied the opinion of people, sentiments, attitudes, assessment, evaluation, and emotions against entities such as products, services, events, organizations, individuals, issues, topics and attributes. This information may be used for market research, product feedback and analyze the effectiveness of customer service. The problem in this research is the selection of features to improve the accuracy of Support Vector Machine and find the parameter values to get the highest accuracy in sentiment analysis review of the delivery of goods as well as produce a classification of negative and positive results of the review appropriately. Author using Principal Component Analysis and Genetic Algorithm as an optimization to improve the accuracy of Support Vector Machine method. The resulting accuracy of the algorithm Support Vector Machine by 86.00%, after optimized by using Principal Component Analysis and Genetic Algorithm accuracy has increased to 97.00%.Keywords: Sentiment Analysis, Review of Goods Delivery, Support Vector Machine.Abstrak - Analisis sentimen mempelajari pendapat orang, sentimen, sikap, penilaian, evaluasi, dan emosi terhadap entitas seperti produk, layanan, kejadian, organisasi, individu, isu, topik dan atribut. Informasi ini dapat digunakan untuk riset pasar, umpan balik produk dan analisis efektivitas layanan pelanggan. Masalah dalam penelitian ini adalah pemilihan fitur untuk meningkatkan akurasi Support Vector Machine dan menemukan nilai parameter untuk mendapatkan akurasi tertinggi dalam analisis sentimen tinjauan terhadap pengiriman barang serta menghasilkan klasifikasi hasil negatif dan positif dari tinjau dengan tepat Penulis menggunakan Principal Component Analysis dan Genetic Algorithm sebagai optimasi untuk meningkatkan akurasi metode Support Vector Machine. Keakuratan yang dihasilkan dari algoritma Support Vector Machine sebesar 86.00%, setelah dioptimalkan dengan menggunakan Principal Component Analysis dan Genetic Algorithm accuracy telah meningkat menjadi 97%.Kata Kunci: Analisis Sentimen, Review Pengiriman Barang, Support Vector Machine.