Automatic resource compilation by analyzing hyperlink structure and associated text

Abstract

We describe the design, prototyping and evaluation of ARC, a

Authoritative Sources in a Hyperlinked Environment

“…In Section 8 we mentioned systems for compiling high-quality www resource lists that have been built using extensions to the algorithms developed here; see Bharat and Henzinger [6] and Chakrabarti et al [10,11]. The implementation of the Bharat-Henzinger system made use of the recently developed Connectivity Server (Bharat et al [5]), which provides very efficient retrieval for linkage information contained in the AltaVista index.…”

“…Since the appearance of the conference version of this paper, three distinct user studies performed by two different groups [6,10,11] have helped assess the value of our technique in the context of a tool for locating information on the www. Each of these studies used a system built primarily on top of the basic algorithm described here, for locating hubs and authorities in a subgraph G σ via the methods discussed in Sections 2 and 3.…”

“…In such cases, the process of diffusion can also provide a broad-topic summary of more prominent pages related to p. Consider, for example, the results when p was sigact.acm.org, the home page of the ACM Special Interest Group on Algorithms and Computation Theory, which focuses on theoretical computer science. The problem of returning more specific answers in the presence of this phenomenon is the subject of on-going work; in Sections 8 and 9, we briefly discuss current work on the use of textual content for the purpose of focusing our approach to link-based analysis [6,10,11]. The use of non-principal eigenvectors, combined with basic term-matching, can be a simple way to extract collections of authoritative pages that are more relevant to a specific query topic.…”

Web Search via Hub Synthesis

“…To get a feel for why this case is easier (when À, and Ë À are all rank ), imagine that the algorithm has available to it the model matrices Ï À Ì and Ë À Ë Ì À , and the actual query Õ Ú Ì Ë Ì À instead of their instantiations 9 . The 8 Another equally easy special case is when À , however, as discussed below, there is empirical evidence suggesting that this is unlikely in practice.…”

“…Kleinberg's seminal paper [20] on hubs and authorities introduced a natural paradigm for classifying and ranking web pages, setting off an avalanche of subsequent work [7,8,10,15,22,9,3,12,19,2,5,27,1,13]. Kleinberg's ideas were implemented in HITS as part of the CLEVER project [7,10].…”