Beydoun, G., Kultchitsky, R., & Manasseh, G. (2007). Evolving semantic web with social navigation. Expert Systems with Applications, 32(2), 265. (edit)

The authors, researchers from the University of Wollongong, Australia and American University of Beirut, believe that the web-surfing experience (history) of user can be beneficiary to other future users of similar interests and should therefore be stored and made reusable. To achieve this, the authors propose the collection of browsing histories (“surfing trails”) and a grouping by category. Instead of categorizing text based search results, the proposed system develops categories based on users from similar community and interests. They identify two steps in the creating of a semantic web: the creating of “smart data” through the use of languages such as XML, and the creating of an ontology to allow interoperability of smart data. However, these steps require Mark-up, which the authors believe to be a process that is cumbersome, errors prone, and time consuming. They therefore proposed a system that only requires users to only label pages visited as good or bad hits, and optionally with a weight. The system uses previous browsing histories to make recommendations by following an embedded reasoning technique. The authors’ system requires four key characteristics: application integration, presence of other users, trust of the source of information, and privacy of the advice giver (similar to Forsberg, Svensson, 1998). Similar to other socialnavigation approaches, this approach is also an incremental one in building meta-data describing the relationships amongst web pages. This approach can be of great interest to users in the research field that require the viewing of several resources at once. The authors propose the use of simple meta-data in order to reduce the load on the user. However this approach may also reduce future possible applications of the meta-data due to its limitation in content.