Association Rules Mining Based Approach for Web Usage Mining

The web has emerged rapidly into a valuable source of information. Web visitors leave trace behind them which is used by web site owners for knowledge discovery. The latter information guides site owners in deciding how to organize the information in their website and how to provide the best for their visitors in order to maximize their profit. Various mining techniques combined with machine learning models could be employed for effective knowledge discovery. The work described in this paper utilizes association rules mining integrated with fuzziness factor in order to analyze weblog data. The target is to find pages that are accessed together by majority of the users and hence should be linked in a proper way in order to maximize user satisfaction by providing to the users the access flow they expect. This way the number of visitors to the analyzed website will be maximized and hence the target will be achieved. Existing systems that are currently in use, such as AxisLogMiner and WebMiner, will be analyzed.