ABSTRACT
The initial part of the book focused on the evolution of web mining from the traditional data mining paradigm. Now it is time for the real application, and the case study covered in this chapter is classification of websites on the basis of standard metrics in order to help the business houses to have Search Engine Optimized (SEO) sticky websites in place. With the growing realm of electronic transactions, the need of such websites is undisputed. The dataset aspects for these websites have already been placed in Chapter 3. The attributes taken for classification are like accessibility, design, texts, multimedia, and networking. These attributes helps in predicting the SEO compatibility of the websites. The best part of this chapter is the actual screen shot and code written in Weka for the exercise of classification using different algorithms, namely J48 (Decision tree-based algorithm), RBFNetwork (neural network based), NaiveBayes (statistical method) and SMO (Support vector machine based). A thorough analysis of the classification vis-à-vis performance of the algorithm is dealt in depth. The know-how carved out of this chapter is not only significant from the web mining point of view, but it is vital in the web usability analysis.
