Abstract
This chapter introduces a personalized ranking function as a means of offering Web information seekers with search results that satisfy their particular interests. It argues that users’ preferences can be accurately identified based on the semantic analysis of their previous searches and that learnt user preferences can be fruitfully employed for personalizing search results. In this respect, we introduce a ranking formula that encapsulates the user’s interests in the process of ordering retrieved results so as to meet the user’s needs. For carrying out our study we relied on a lexical ontology that encodes a number of concepts and their interrelations and which helps us determine the semantics of both the query keywords and the query matching pages. Based on the correlation between the query and document semantics, our model decides upon the ordering of search results so that these are personalized.
TopBackground
There has been previous work in personalizing Web search. One approach to personalization is to have users explicitly describe their general search interests, which are stored as personal profiles (Pazzani, Muramatsu, & Billsus, 1996). Personal profiles, specified explicitly by the users have also been used to personalize rankings, such as the PageRank algorithm (Aktas, Nacar, & Menczer, 2004; Jeh & Widom, 2003). There also exist many works on the automatic learning of a user’s preference based on the analysis of the user’s past clickthrough history (Chen & Sycara, 2004; Pretschner & Gauch, 1999, Sugiyama, Hatano, & Yoshikawa, 2004). Pretschner and Gauch (1999) for instance, describe how a user’s preference is identified based on the five most frequent topics in the user’s log data.
On the other hand, Chen and Sycara (2004) generate multiple TF-IDF vectors, each representing the user’s interests in one area. Sugiyama et al. (2004) employ collaborative filtering techniques for learning the user’s preference from both the pages the user visited and those visited by users with similar interests. Likewise, Teevan, Dumais, and Horvitz (2005) employ rich models of user interests, built from both search-related information and information about the documents a user has read, created, and/or e-mailed.
Key Terms in this Chapter
Search Engine: A search engine is an information retrieval system designed to help find information stored on a computer system, such as on the World Wide Web, inside a corporate or proprietary network, or in a personal computer.
PageRank: A link analysis algorithm which assigns a numerical weighting to each element of a hyperlinked set of documents, such as the World Wide Web, with the purpose of “measuring” its relative importance within the set. The algorithm may be applied to any collection of entities with reciprocal quotations and references. The numerical weight that it assigns to any given element E is also called the PageRank of E and denoted by PR(E).
Open Directory: The Open Directory Project (ODP) is a multilingual open content directory of World Wide Web links that is constructed and maintained by a community of volunteer editors.
Topi cal Ontology: In both computer science and information science, an ontology is a data model that represents a set of concepts within a domain and the relationships between those concepts. It is used to reason about the objects within that domain.
Collaborative Filtering: The method of making automatic predictions (filtering) about the interests of a user by collecting taste information from many users (collaborating). The underlying assumption of collaborative filtering approach is that those who agreed in the past tend to agree again in the future.
WordNet: WordNet is a semantic lexicon for the English language. It groups English words into sets of synonyms called synsets, provides short, general definitions, and records the various semantic relations between these synonym sets. The purpose is twofold: to produce a combination of dictionary and thesaurus that is more intuitively usable, and to support automatic text analysis and artificial intelligence applications.
Search Personalization: Search results customization for specific interests is widely known as personalized search. Personalized search has a significant potential in providing users with information that greatly satisfies their particular search intentions.