A Web Surfing Assistant for Improved Web Accessibility

Due to the exponential increase of information, search and access for the Web information or service takes much time. Web information is represented through several web pages using hyperlinks and each web page is contains several topics. However. most existing web tools don't reflect such web authoring tendencies and treat it as an independent information unit. This inconsistency yields inherent problems in web browsing and searching. In this paper, we propose a web surfing assistant called LinkBroker that provides collodion pages. They are composed of relevant information extracted from several web pages that have table and frame structure in order to improve accessibility to web information. Especially, the system extracts a set of web pages that are logically connected and groups those pages using table and frame tags. Then, essential information blocks in each page of a group are extracted to construct an integrated summary page. It Provides a comprehensive view to user and one cut way to access distributed information. Experimental results show the effectiveness and usefulness of LinkBroker system.