Authors : Harwati , Agus Mansur and Adnan Karunia
Abstract: Social network has grown rapidly in line with the development of technology. Twitter is one of social media that has worldwide. It is widely used among individual user as well as companies or organizations to communicate with their consumer. Consumer’s tweet can be utilized by companies to determine consumer behavior in response to their product or service. This method is more efficient than mining voice of by distributing questionnaires. Comments written by customers is usually tend to original and not contrived. The objective of this research is to map the consumer behaviors based on the opinion in Twitter. Smartphone iPhone 5 as a trending topic smartphone in Twitter is taken as the object for this research. Clustering text mining techniques are used to settle the problem. Two level clustering is done to get the mapping of consumer behavior. In the first step, RapidMiner is used to cluster more 200 comments in to 8 groups. Second step is called by profiling groups. It obtains three clusters by analyzing the similarity characteristics and the meaning of the comments contained in each group. The resulting final cluster are cluster of 29.20% positive comments, cluster of 42% negative comments and a cluster of other comments by 28.80%. From this result, it can be concluded that based on their comments in the social media Twitter, the unsatisfied consumers is greater than others.
Harwati , Agus Mansur and Adnan Karunia, 2017. Utilization of Social Media for Consumer Behavior Clustering using Text Mining Method. Journal of Engineering and Applied Sciences, 12: 6406-6411.