A neuro-fuzzy classifier for website quality prediction

To improve the quality of websites, it is necessary to continually assess and evaluate web metrics and subsequently make improvements. In this research, we have computed nine quantitative web measures for each website using an automated Web Metrics Analyzer tool developed in JAVA programming language. The website quality prediction models are developed utilizing ANFIS-Subtractive clustering and ANFIS-FCM based FIS models, to classify the quality of website as good or bad. The models are validated using 10 cross validation on a collection of web pages of Pixel Awards web metrics collected through the tool. The results are analyzed using Area Under Curve obtained from Receiver Operating Characteristic (ROC) analysis. The results showed that both ANFIS-Subtractive and ANFIS-FCM have acceptable performance in terms of specificity and sensitivity. In addition, ANFIS-Subtractive and ANFIS-FCM clearly induces only two rules, which are much less than 512 rules generated by the normal ANFIS model. Hence ANFIS-Subtractive and ANFIS-FCM are the most comprehensible techniques tested in this work.