Provide a warning system for critical situations in market research using fuzzy neural network
Saman Siadati *, M. Jafar Tarokh **
Consumers rely more and more on online comments and ideas on the Internet when making purchasing decisions. Hence, companies should be aware of the true position of their products on the Web. For this reason, in this paper, a warning system is proposed for online market research that allows identifying critical situations when online thoughts are formed. In this way, the system sends an alert to the marketing manager and allows marketers to take precautionary measures. The alert system operates based on a knowledge base that contains product success rates, online comments, and social interaction patterns. This knowledge comes from using techniques such as text mining and social networking analysis. Based on this knowledge, the warning system decides and judges in the circumstances. To this end, a fuzzy approach is chosen that learns the rules of language from the data. These rules are used to estimate future positions. In order to evaluate this system, two scenarios have been used, which shows that all components of the warning system have better performance than alternative methods.
Warning system, Online product research, Opinion Mining, Social network analysis, Fuzzy-Neural System
*,** Strategic Intelligence Research Laboratory, K.N.T. university of technology, Tehran, Iran