S²ProT: rank allocation by superpositioned propagation of topic relevance

Abstract

Purpose-To assign topic-specific ranks to web pages. Methodology/Approach-The main idea of our approach is to use power iteration to assign topic-specific rating values (called relevance) to web pages, creating a ranking or partial order among these pages for each topic. Our approach depends on 1) a set of pages that are initially assumed to be relevant for a specific topic, 2) the spatial link structure of the web pages, and 3) a net-specific decay factor that we call ξ. Findings-Our approach exhibits desirable properties such as fast convergence, stability and yields relevant answer sets. The first property will be shown using theoretical proofs, while the others are evaluated through stability experiments and assessments of real world data in comparison with already established algorithms. Research limitations/implications-In our assessment, we used all pages that our web spider was able to find in the Nordic countries. It is also important to note that entities that use domains outside the Nordic countries (e.g., .com or .org) are not present in our datasets even though they reside logically within one or more of the Nordic countries. This is quite a large dataset, but still small in comparison with the entire World Wide Web. Moreover, the execution speed of some of the algorithms unfortunately prohibited the use of our large test dataset in the stability tests. Practical implications-It is not only possible, but also reasonable, to perform ranking of web pages without using Markov chain approaches. This means that the work of generating answer sets for complex questions could (at least in theory) be divided into smaller parts that are later summed up to give the final answer. Originality/value-This paper contributes to the research on Internet Search Engines.