Analyzing the Correlation Between Tweets and Sales for Product Brands

Abstract

In recent years, Social Networking Services (SNS) have become widespread, and anyone can freely spread information via this Consumer Generated Media (CGM).The Social Media Data generated in such a way contains a great deal of opinions and values of consumers, and active utilization of this data is being considered. In this paper, in order to develop a method to obtain more informative information about consumer purchasing behavior, we targeted the carbonated drink brand "Product Brand A", and analyzed the correlation between tweets including a product brand’s name and sales to capture the correlation; then we identified and excluded information(noise) other than information about consumer purchasing behavior included in the tweets information for a more detailed correlation analysis. In addition, by applying the analysis of the qualitative aspects of tweets(emotion evaluation analysis) to actual data and analyzing them, we suggested that further possibility of data utilization.