Current search engines require the explicit specification of queries in retrieving related materials. Based on personalized information acquired over time, such retrieval systems aggregate or approximate the intent of users. In this case, an aggregated user profile is often constructed, with minimal application of contextspecific information. This paper describes the design and realization of the idea of 'Links into the Future' for discovering related documents from the Web, within the context of an electronic journal. The information captured based on an individual's current activity is applied for discovering relevant information along a temporal domain. This information is further pushed directly to the users' local contexts. This paper as such presents a framework for the characterization and discovery of highly relevant documents.1