PageRank without hyperlinks: Structural re-ranking using links induced by language models

TL;DR

This paper introduces a novel re-ranking method for information retrieval that leverages language model-induced generation links between documents, improving precision at top ranks without bias against long documents.

Contribution

It proposes a structural re-ranking approach using generation links derived from language models, enhancing retrieval effectiveness beyond traditional methods.

Findings

Centrality-based re-ranking improves top-rank precision.

Generation links effectively capture document relationships.

Method reduces bias against long documents.

Abstract

Inspired by the PageRank and HITS (hubs and authorities) algorithms for Web search, we propose a structural re-ranking approach to ad hoc information retrieval: we reorder the documents in an initially retrieved set by exploiting asymmetric relationships between them. Specifically, we consider generation links, which indicate that the language model induced from one document assigns high probability to the text of another; in doing so, we take care to prevent bias against long documents. We study a number of re-ranking criteria based on measures of centrality in the graphs formed by generation links, and show that integrating centrality into standard language-model-based retrieval is quite effective at improving precision at top ranks.