The Web is a vast playground for the propagation of information by individuals. The capability of Web users to leverage the value of Web content, in terms of factors such as usefulness, reputation, or reliability, has emerged as a requirement. The most significant characteristics of the Web are hyperlinks, which enables an author to refer to Web content published by other people. As a result, Web content can refer to other content on the Web in various contexts. The references provide important clues to understanding the roles or reputations of various Web pages according to the viewpoints of third parties. In this paper we propose an approach to mining referential contexts on the Web.