This paper examines how to use the concept of entropy to discover rare and unexpected patterns in websites. First, by collecting and pre-processing the log data of ehadish.com website, the required data is prepared for analysis. Then, using the concept of entropy as a measure to measure discontinuity and fluctuation in data, rare and unexpected patterns in the behavior of website users and customers are identified. In the next step, using the k-medoids clustering method, website users are divided into similar groups and the frequency patterns of visits and customer behaviors in each group are examined. Then, using the monitoring technique, suspicious patterns and sudden changes in the website are identified and evaluated, and entropy is used as an important tool to detect these changes. Finally, by producing a management dashboard including reports and analytical results in a graphical and orderly manner, the obtained information is presented to the website administrators. This dashboard allows administrators to compare the website's performance and security with similar examples and take necessary actions to improve their website's security and performance. As an analytical and preventive method, this article can help improve the security and quality of services provided by websites and provide managers with valuable information about user behavior and customer patterns.