Citation Context

... from user profiling techniques. Profile-based methods can be potentially effective for almost all sorts of queries, but are reported to be unstable under some circumstances [1]. Although there are pros and cons for both types of PWS techniques, the profile-based PWS has demonstrated more effectiveness in improving the quality of web search recently, with increasing usage of personal and behavior information to profile its users, which is usually gathered implicitly from query history [2], [3], [4], browsing history [5], [6], click-through data [7], [8], [1] bookmarks [9], user documents [2], [10], and so forth. Unfortunately, such implicitly collected personal data can easily reveal a gamut of user’s private life. Privacy issues rising from the lack of protection for such data, for instance the AOL query logs scandal [11], not only raise panic among individual users, but also dampen the data-publisher’s enthusiasm in offering personalized service. In fact, privacy concerns have become the major barrier for wide proliferation of PWS services. 1.1 Motivations To protect user privacy in profile-based PWS, researchers have to consider two contradicting effects during the search process. O...

Citation Context

...l data, which never leaves the user’s device. Shifting the behavioral targeting computations to the client device is not a new idea and has been recently applied to the context of personalized search =-=[16]-=- and online advertising [12, 7, 6, 3]. However, targeted coupon delivery poses additional challenges beyond those demanded by targeted advertising. Shifting the computations for behavioral matching to...

Citation Context

...he point of view of individuals over the data usage. An204 S.-E. Tbahriti et al. approach on the feasibility of achieving a balance between consumers privacy and provider search has been proposed in =-=[21]-=-. It allows client to collect, summarize, and organize their personal information into a hierarchical profile. Through this profile, the client controls which portion of its private information is exp...

Citation Context

... individual users, such as information captured by the index of desktop search services, and used only in privately-held, local analyses, can be used in order to provide personalization of web search =-=[23, 25]-=-. These techniques have demonstrated the potential of greatly improving the relevance of displayed search results by disambiguating queries based on personal information about the users. On the implic...

Citation Context

... Page rank is based on number of times pagesclicked for information. Average ranking is nothing butscalculating average of number that times the search is given.sY. Xu, K. Wang, B. Zhang, and Z. Chen =-=[12]-=- contributes insenhancing privacy in personalize web search. To access thesuser profile full access to server is required. They show thatsmany servers access this user information without permissionso...

Citation Context

...e retrieval, we take some preprocessing steps like stop words removal and stemming. In fact, each case is treated as a list of terms after preprocessing. Constructing a Hierarchical User Profile In ExpertRec, any users’ experiences, i.e. search histories and favorites could be organized into search cases and the data source for user profiles. Our hypothesis is that terms that frequently appear in snippets, titles, queries and tags of search cases can represent topics that interest users, and votes and selected-frequency of search cases influence degree of interests. Here we use an approach in [14] proposed by Xu et al. to build the hierarchical user profile based on frequent terms. In the hierarchy, general terms with higher frequency are placed at higher levels, and specific terms with lower frequency are placed at lower levels. C represents the collection of all search cases in one ExpBase. C(t) denotes all cases covered by term t, i.e. all cases in which t appears, and |C(t) |represents the number of cases covered by t. A term t is frequent if |C(t) |≥ minsup, where minsup is a system-specified threshold, which represents the minimum number of cases in which a frequent term is requi...

Citation Context

...ion IV and Conclusion and Future Work in section V. II. RELATED WORK Personalized web search is mainly focused on better data utility and privacy. We proposed user’s profile protection with this framework and efficiency of search results is increased with re-ranking technique [14], [15], [16]. The most of the search engines use re-ranking technique for image retrieval. The user’s profile generation of this framework is in hierarchical structure with weighted graph such as ODP [2], [17], [18]. The performance of the personalized web search is estimated with metrics like average precision [19], [20], rank scoring [21] and average rank [4], [22]. Identity disclosure is described in [9]. In this paper we have analysed the threat with privacy preserving data publishing [13] algorithms. ISSN(Online): 2320-9801 ISSN (Print): 2320-9798 International Journal of Innovative Research in Computer and Communication Engineering (An ISO 3297: 2007 Certified Organization) Vol. 3, Issue 4, April 2015 Copyright to IJIRCCE 10.15680/ijircce.2015.0304073 3400 III. PERSONALIZED WEB SEARCH FRAMEWORK The personalized web search is a framework where the user profile is protected during the search. The PWS is us...

Citation Context

...be protected, which are hard to define in the web context in practice, or only guarantee the anonymity of a user but do not prevent the linking attack between a user and a potentially sensitive item. =-=[23]-=- proposed a technique for building user profiles with configurable levels of details. A few recent works specifically studied anonymizing query logs. Notably, [13, 11] have demonstrated the ineffectiv...