A Hybrid Framework for Building a Web-Page Recommender System

Author

Anastopoulos, Vasileios ; Karampelas, Panagiotis ; Kalagiakos, Panagiotis ; Alhajj, Reda

Author_Institution

Hellenic American Univ., NH, USA

fYear

2011

fDate

12-14 Sept. 2011

Firstpage

385

Lastpage

390

Abstract

Recommender systems aim to facilitate World Wide Web users against information and product overloading. They are usually intermediate programs that try to predict users´ preferences and items of their interest. In this paper, we present a hybrid framework that uses open source information such as web logs in combination with social network analysis and data mining, to extract useful information about users browsing patterns and construct a recommendation engine. A case study based on real data from an organization of 250 employees is presented and a system prototype is constructed based on the results.

Keywords

Internet; data mining; information retrieval; recommender systems; social networking (online); Web logs; Web-page recommender system; World Wide Web; data mining; information extraction; information overloading; open source information; product overloading; recommendation engine; social network analysis; Association rules; IP networks; Itemsets; Recommender systems; Social network services; Web pages; association rules; data mining; hybrid framework; recommender system; social network; system prototype;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligence and Security Informatics Conference (EISIC), 2011 European

Conference_Location

Athens

Print_ISBN

978-1-4577-1464-1

Electronic_ISBN

978-0-7695-4406-9

Type

conf

DOI

10.1109/EISIC.2011.40

Filename

6061270