In this paper we describe a method for identification of a particular user in large information spaces such as the Web is. Our main goal is to be able to decide whether the particular information found on the Web is relevant to the person being looked-up or not by taking into account additional background knowledge about that person: his or her social network. Our method combines semantically as well as syntactically based metrics to compare different social networks acquired during the identification process. We describe evaluation of the proposed method and its comparison to the related works.