Cosmetic Product Review Sentiment Using K-Nearest Neighbor

Abstract

Indonesia, with a population of about 250 million, is a promising market for cosmetics companies. Aiming for women as the main target group for consumers, most of the cosmetics industry has recently begun to innovate products for men (Ministry of Industry, Republic of Indonesia, 2013). According to the Indonesian Cosmetics Companies Association (Perkosmi) of the Ministry of Industry of the Republic of Indonesia, sales of imported cosmetics reached IDR 2.44 trillion in 2012, an increase of 30% compared to IDR 1.87 trillion in 2011. In 2013, sales of imported cosmetics are expected to increase by another 30% to Rs 3.17 trillion. One of the most popular cosmetics is the MS Glow brand. MS Glow is an acronym for Magic For Skin, the brand's motto, which reflects Indonesia's most brilliant products. The data is from WebFemaleDaily, where users review Indonesian MS Glow cosmetics. The results of the analysis are intended to determine the positive and negative reactions of users of MS Glow cosmetics. The classification process uses the K-nearest neighbor (K-NN) algorithm. The accuracy value of K-NN with k = 1 is obtained from the result of 63.64% confusion matrix test.