Category ranking for personalized search

Supporting personalized ranking over categorical attributes

“…In particular, this work pre-computes a set of 16 rank vectors for each of the 16 top categories in ''Open Directory Project (ODP)", which are combined to generate a personalized ranking for a specific user query. Recently, more efforts followed to use ontologies [20] or ODP categories [14,6,15] to enhance retrieval. More specifically, Ref.…”

“…Similarly, Ref. [15] models user profile as a ''graph" of categories, which is obtained from implicit user browsing behaviors. In addition, personalization techniques have also been widely adopted to other domains, such as personalized web advertising [12] or personalized recommendation [22].…”

Concept-Based, Personalized Web Information Gathering: A Survey

“…In order to acquire user profiles, Chirita et al [6] and Teevan et al [57] extracted user interests from the collection of user desktop information such as text documents, emails, and cached Web pages. Makris et al [37] comprised user profiles by a ranked local set of categories and then utilized Web pages to personalize search results for a user. These non-interviewing techniques, however, have a common limitation of ineffectiveness.…”

A User Profiles Acquiring Approach Using Pseudo-Relevance Feedback

“…text documents, emails, and cached Web pages. Makris et al [16] comprised user profiles by a ranked local set of categories and then utilized Web pages to personalize search results for users. These works attempted to acquire user profiles by discovering user background knowledge first.…”

“…Such user profiles are usually acquired by observing and mining knowledge from users' activity and behavior [25]. Typical models are [6] and [20]'s ontological user profiles, and also models developed by [8,14,16]. They acquired user profiles adaptively based on the content of user queries and online browsing history.…”