Customer data mining with RF-T algorithm

Roya Mahmoudi¹ Hassan Shakeri² Mansoureh Zare³

1) Ph.D. student e-commerce, Department of Information Technology and Computer Engineering, Sabzevar Branch, Islamic Azad University of Sabzevar, Iran. Email:
2) Assistant Professor, Department of Computer, Islamic Azad University, Mashhad Branch, Iran -
3) Ph.D. student e-commerce, Department of Information Technology and Computer Engineering, Sabzevar Branch, Islamic Azad University of Sabzevar, Iran. Email:

Publication : 2nd. International Congress on science & Engineering - paris(parisconf.com)
Abstract :
By examining customer data and reaching an understanding of customer demand, customer relationships can be enhanced. Therefore, achieving effective methods in analyzing customer data, especially in today s organizations, is very fundamental. The use of data mining methods to predict and identify customer needs has been widely studied from different approaches. Text data mining is one of the most important approaches to data analysis, which analysis of this valuable data can provide a better understanding of customer behavior and the appropriate response to it in the organization. Text mining using random forest algorithms can help us in data mining using textual data. The forest algorithm is random for categorization, which is used to classify the semantic data extracted from the text. These classifications will help us anticipate customer needs. In this study, an attempt has been made to examine the random forest algorithm for extracting keywords from customer data. The results are then compared with other classification algorithms, which show that the random forest performs better than other classifications and has the best results in classifying the bigram features extracted from the text.
Keywords : Data Mining Text Mining Customer Random Forest.