Web Site Visit Forecasting Using Data Mining Techniques
Chandana Napagoda · International journal of scientific and technology research · 2013
Abstract : Data mining is a technique which is used for identifying relationships between various large amounts of data in many areas including scientific research, business planning, traffic analysis, clinical trial data mining etc. This research will be researching applicability of data mining techniques in web site visit prediction domain. Here we will be concentrating on time series regression techniques which will be used to analyse and forecast time dependent data points. Then how those techniques will be applied to forecast web site visits will be explained. Keywords : Forecasting, Web Site, SMO Regression, Linear Regression, Gaussian Regression and Multilayer Perceptron ———————————————————— I. I NTRODUCTION In contemporary world with the technological changes, ability of predicting web site visiting patterns have a significance value for every site owner for targeting their business for right customers at right time. The available solutions to predict future access and usage patterns for web sites are Business Intelligence (BI) tools. But for many small and medium sized companies or web site owners are unable to afford for them since they are expensive solutions. The available BI tools including Google Analytic doesn't support forecasting features. They are focused on analysing user behaviours on the web sites and log related results. So the solution going to be developed from this research is targeted on web site owners and administrators to assist the future predictions of web site visits on their marketing strategies. For instance in software products related companies, they need to know the visiting history for their sites to plan their new releases, upgrades, etc. Another case is density of forecast visitors would be helpful to allocate or deallocates servers. All the above requirements exist in web site visit prediction domain motivated me on this research work for providing a suitable solution implementation. The aim of this research work is applying a suitable forecasting technique to predict web site visits. Thus, the results derived through forecasting will be assist web site owners in,