Identification of Karleen Hijab Fashion SME Competitors Based on Sentiment Analysis Using Naïve Bayes Classifier Algorithm

Abstract

Hijab Fashion Small and Medium-Sized Enterprises (SMEs) need to develop competitive advantages brands as a source of SME competitiveness. However, most Hijab Fashion SMEs experience limitations in developing the competitive advantages of their brands. This research was conducted to find out and understand the competitive advantages of Karleen Hijab Fashion SME competitors as the object of study. The method used is sentiment analysis using the Naïve Bayes algorithm. Sentiment analysis was carried out using online review data of Shopee e-commerce. Sentiment analysis data processing was done using orange data mining software. Sentiment analysis using the Naïve Bayes algorithm produced an average value of AUC, CA, F1, Precision and adequate recall for the entire Hijab Fashion SME brand, which is 0.72, 0.887, 0.856, 0.833, and 0.887. Based on the percentage of the largest positive sentiment on each fashion quality attribute, it is known that competitive advantages of Lozy are in the Fabric Quality Attribute (30.77%), and Good Fit (15.38%), and Halwa's competitive advantage is in the Design attribute (34.19%).  Competitive advantages of Hijup are on the Serviceability Attribute (21.74%) and Packaging (15.38%), and Competitive advantages of Lafiye are on the Price Attribute (6.17%). Competitive advantages of Deenay brand are on the Reliability Attribute (20.89%), while Karleen does not have a relative advantage on any fashion quality attribute because the percentage of positive sentiment for each attribute is still below competitors.