Over the years there have been many attacks of Phishing and many people have lost huge sums of money by becoming a victim of phishing attack. In a phishing attack emails are sent to user claiming to be a legitimate organization, where in the email asks user to enter information like name, telephone, bank account number important passwords etc. such emails direct the user to a website where in user enters this personal information. These websites also known as phishing website now steal the entered user information and carries out illegal transactions thus causing harm to the user. Phishing website and their mails are sent to millions of users daily and thus are still a big concern for cyber security. Social engineering has come up with many educational and training programs to make users be aware of phishing website and avoid users to become victim of such attacks. Usually a phishing website can be easily identified by its URL, its email links or HTML code. Thus many automatic phishing classifiers are being built to classify whether the given mail or website is a phishing website or not. Data mining techniques, Machine algorithms techniques and programming can help in developing a system capable enough to classify whether a website is a phishing website or not. In this research work I use the dataset of phishing website of UCI machine learning dataset and data mining concepts to understand the pattern of phishing website. I select some classifiers compare their results over the given dataset and select among them the best classifier to make a machine learning base phishing website detection system. I make use of R Script interfaced with WEKA 3.2 to help in detecting phishing website.