An Approach To Build Sequence Database From Web Log Data For Webpage Access Prediction

Sequential data mining is one of important topics in data mining. One of its important application is to predict a next element in a data sequence or to discover sequential rules. Many algorithms were published to tackle these problems such as sequential rule mining and sequence prediction. This paper aims to present an approach to build a sequence database from a Web log data. This step accounts for first phase of sequential data mining, titled data pre-processing phase, which transforms sequential data in Web log into sequential relational database. This sequential relational database is used as the input data of sequential data mining and sequence prediction. Besides, we present a parallel algorithm to efficiently build sequential database from Weblog files in order to predict Webpage access.