Phishing attacks allure users to submit their personal information to fake websites that mimic legitimate websites. Many anti-phishing techniques have emerged in recent years. However, the numbers of phishing attacks are still increasing. Two reasons can be blamed for this situation. First, users have too much trust and confidence on existing anti-phishing tools in general. Second, most users believe that they are foolproof against phishing attacks when anti-phishing tools are deployed. We believe that understanding of anti-phishing tools based on their common features can be the beginning step to address these issues. However, there is no extensive analysis of existing anti-phishing techniques. This paper attempts to classify existing works based on information sources. The classification would not only provide useful information to develop new anti-phishing techniques or improve existing techniques, but also enable our understanding on the limitations of the existing techniques.