Article Preview
TopIntroduction
Customer/User segmentation has been a critical element of the marketing and is one of the most important strategic concepts contributed by the marketing discipline to business firms (Shina et al., 2004). User Segmentation is the process of developing effective schemes for categorizing and organizing meaningful groups of customers. A user segment is a group of prospects or customers who are selected from a database based on characteristics they possess or exhibit. It also allows company to differentially treat consumers in different segments. User Profiling is the process of analyzing the customers of each segment in order to generalize, describe or name this set of customers based on common characteristics. It is the process of understanding and labeling a set of users. It provides valuable information about users/customers so marketers can furnish stronger, more targeted offers and each user segment is the target group of users.
One-to-one marketing is the ideal marketing strategy, in which every marketing campaign or product is optimally targeted for each individual customer; but this is not always possible. Therefore, segmentation is required to distinguish similar users and put them together in a segment/group. Doubtlessly using segmentation to understand user’s needs is much easier, faster and more economical than uniquely investing to understand them particularly (Jonker et al., 2004).
With proper market segmentation, enterprises can arrange the right web pages, services and resources to each target user group and build a close relationship with them. Market segmentation has consequently been regarded as one of the most critical elements in achieving successful modern marketing and customer relationship management.
Click stream data is a sequence of Uniform Resource Locators (URLs) browsed by the user within a particular period of time. To discover group of users with similar behavior and motivation for visiting the particular website can be found by clustering. Traditional clustering (Lee et al., 2008) is used to segment the web users or web pages in to groups based on the existing similarities. When a clustering method is used for grouping users, it typically partitions users according to their similarity of browsing behavior under all pages. But, in the most cases web users behave similarly only on a subset of pages and their behavior is not similar over the rest of the pages. Therefore, traditional clustering methods fail to identify such user groups.
To overcome this problem, concept of Biclustering was introduced. Biclustering (Bleuler et al., 2004; Chakraborty et al., 2005; Madeira et al., 2004) was first introduced by Hartigan (1972). Biclustering is the simultaneous clustering of rows and columns of the data matrix. In literature, biclustering algorithms are widely applied to the gene expression data. In this paper, it is used to mine clickstream data in order to extract target usage groups. These groups are analyzed to determine user’s behavior which is an important element in the E-Commerce applications.
The rest of the paper is organized as follows. In Section 2, some of the existing work related to the biclustering approaches and user segmentations are discussed. Methods and materials required for biclustering approach are described in Section 3. Section 4 focuses on the proposed Biclustering approach using Genetic Algorithm. Analysis of experimental results is discussed in Section 5. Section 6 concludes the paper.