Opinion mining for skin care products on twitter

Abstract

Nowadays, the popularity in using skin care tends to increase and there are also a lot of exchanging opinions on online media, which directly affected to making decision on buying any products for customers. In this research, we want to find additional data for developing opinion analysis and separating emotional opinions about skin care messages. The methodology uses the data mining process, such as opinion mining with sentiment analysis through the machine learning algorithm for identifying the levels of positive and negative emotion in messages. Moreover, the skin care opinion mining application was developed based on the web application to display the results in the form of various representations. Furthermore, the performance of analytical methods is evaluated by the accuracy, precision, and recall rate, which are all more than 75%. Therefore, the automate analysis application can be employed as a helping tool for data analysis for the consumers, who are interested in skin care products, and for the entrepreneurs can know the customers’ attitude of the products.