Internet Usage Pattern

ISPs always confront a huge amount of their customer usage raw data. Since these data are too voluminous, data mining techniques can be useful to recognize hidden usage patterns of the users to develop relevant management strategies. But these patterns vary according to variable circumstances during a year. Therefore, in this paper, the usage patterns of internet users are detected continuously during time line applying usage attributes of the users and association rules of mining technique within an incremental data mining framework. Raw data of an ISP is surveyed to describe the model. The results of the article show that the usage patterns of internet users obtained from this data mining model can lead to the improvement of ISP services and the reduction of their costs.