There are various types of approaches to user profile acquisition, classified into five groups: (1) data mining, (2) statistics and network analysis, (3) retrieval of information, (4) Machine learning and (5) Cognitive. Most of the methods concern static websites, except a couple of methods that can be applied to dynamic websites (Nasraoui & Rojas, 2003). The method uses data mining techniques such as frequent patterns and reference mining found by (Holland et al. , 2003; KieBling & Kostler, 2002) and (Ivancy & Vajk, 2006). Frequent and reference mining is an intensive research area in data mining with a wide range of applications to discover a pattern from web log data to gain insights into users' browsing behavior. The frequent and reference patterns of their search can be classified into page sets, page sequences and page graphs. The use of descriptive statistics to extract knowledge from Web logs was introduced by Srivastava, Deshpende & Phang (2000), analyzing session files and performing user interaction statistics such as frequency, mean and median on variables such as views of page, viewing time and length of a navigation path. Furthermore, the analysis of Web log files using the statistical approach proposed by Stermsek et al. (2007) allows for a broader perception of user behavior and the potential to improve user profiling. Their approach includes several methods such as statistical inference, graphical analysis and profile generation: (1) statistical inference comes from pre-processed web log data, (2) structure analysis to select a certain structure on the website (for example, football news), then carry out the website structure according to the user's interests in football news and (3) Graphic analysis or...... middle of the paper .... .. of Web personalization to personalize services (e.g. content and articles) based on needs and preferences of the user or group of users. This is clear in the adaptive Web, where acceptance means meeting user needs (Agrawal, 1999; Mobasher et.al., 2000). There are several approaches to representing user profiles. For example, Mobasher et al. (1999) and Kang et al. (2001) analyze URLs to obtain user preferences from Web logs. Furthermore, Mobasher et.al. (1999) combined Web usage extraction and content extraction for effective personalization. Shokry and Ajinanth (2006) used clustering techniques to obtain user preferences. From the literature review, the UP representation method can be classified into three methods: (1) graphical representation, (2) tree, (3) XML, and (4) RDF. The UP representation method in Web customization is briefly described below, as presented in the Table 8.