Online advertising is about publishing advertisements/commercials on the Web and helps advertisers to achieve their target on the Web. Online advertising maintains a set of popular websites on their network for each market/country. Therefore, they have to forecast the traffic of these websites. This information will be helpful for business analysts to propose the suitable Web sites to the marketers for advertising their product. The Business analysts have to analyse the user patterns; i.e., traffic data, demographics, etc., of various websites in their network before they propose a deal to the marketers. Most of the traffic on the websites is significantly steady. However, traffic data on few of the websites varies due to some periodic special events (like cricket world cup, rugby world cup, etc.) or sudden cases (like natural disasters) and some are seasonal websites (skiing websites, Christmas, etc.). All these factors have to be considered while forecasting the traffic of the Websites. Thus, online advertising have to predict the traffic of every website depending on their historical traffic data for planning or for scheduling commercials for Clients. Current research mainly concentrates on the data present on World Wide Web (WWW). Employing various data mining schemes to unearth the underlying patters from the web is termed as Web mining. This stream of data mining processes the data that is present in form of web pages or web activities (for ex: server logs) (Dunham, 2003). Web mining tasks can be divided into three types, which are Web usage mining, Web content mining and Web structure mining. This research is primarily concentrating on Web usage mining. Web usage mining mainly involves the automatic discovery of user access patterns from one or more Web servers (Mobasher, 1997). The analysis of such data can help the organizations to determine the life time value of customers, cross marketing strategies across products, and effectiveness of promotional campaigns. Finally, for organizations that sell advertising on the World Wide Web, analysing user access patterns helps in targeting ads to specific groups of users (Mobasher, 1997). Therefore, Web usage patterns can be used to acquire business intelligence to improve sales and advertisement on the Web. The main objectives of this research are to mine four years historical data and identify the hot spots of websites, to discover the intensity and the time span of hot spots, and how these spots can be used in future traffic prediction. In addition, we are interested to research how these hot spots are recurring year after year. Current research mainly studies the historical traffic data of the websites and attempts to predict the future traffic by using the data mining models. Each and every data mining model has certain advantages and disadvantages. Some are suitable to certain domains and some are inappropriate to others. Therefore, the challenge is to determine the suitable data mining model for the current domain.