PROFILE BASED PERSONALIZED WEB SEARCH USING GREEDY ALGORITHMS

B Sekharbabu, P Prasanna, D Rajeswara, J Lakshmianusha, A Pratyusha, A Chand
2016 unpublished
Internet usage is being increased as it provides information to all the users. The required information is retrieved by the search engines. The currently working search engines using sophisticated algorithms will not always provide relevant information to user's requirements. To resolve the issue, Personalized web search is used that will improve the quality of the search result by reordering the search results. This web search is done to provide relevant results using the user profile. The
more » ... osed UPS framework will dynamically generate a user profile for a user's query prioritizing the user's privacy. To acquire this we are implementing Greedy DP and Greedy IL Algorithms that are used for runtime generalization. INTRODUCTION A web search Engine is a software system that enables to search for the information on web (www) [2]. The information mined can be images, documents or any other type of file. The usefulness or performance of the search engine depends on the relevance of results it mines. Some search engines use ranking [5] method to the web pages for better results but the ordering varies from one engine to another. The search Engines mainly can be two types where the search engine creates a list of results by using software called CRAWL, a crawler based search engine like GOOGLE and the other depends on the human editors like YAHOO [3], Google directory generally said to be human powered directories. The directory search is good to reveal relevant results in a different topic search. If the user gives a query named "virus" to the search engine expecting the result of a malware program but the engine might present a result of a parasite or an infectious agent. Based on the perspective of the user the result is not acquired in such situations. The same continues with the query "Kingfisher" that can be a bird or an airlines resulting in occurrence of ambiguity [4]. The web information systems are thus facing many challenges. The data is being increased day by day resulting in information overloading is one of the major challenges. In such situations behavior and the content measures and considered. However all the search engines work with a specific goal of listing the relevant results implementing different algorithms? Web Personalization will resolve some of the issues that were being faced by information systems. Generic search follows "one size fits all" model that cannot retrieve results dynamically like a personalized web search. Main goal is to retrieve the relevant results for any kind of user based on one's interest. The solution to personalized web search (PWS) can be profile based and click-log based. Click log based method is implemented by referring to a user's query history. But this can only work on repeated queries provided by same user which is a limitation. Profile based method is implemented by using user interest model that is generated by user profiling. But
fatcat:ai3cwfgzlffghh2oewaw23zcgy