Biased TextRank: Unsupervised Graph-Based Content Extraction

TL;DR

Biased TextRank is an unsupervised, graph-based content extraction method that improves focused summarization and explanation extraction by incorporating relevance to a specific focus, leading to better performance and efficiency.

Contribution

It introduces a novel biasing mechanism in TextRank through relevance-based random restarts, enhancing focused content extraction.

Findings

Improved ROUGE-N scores on two datasets.

Faster and lighter than current state-of-the-art methods.

Effective for focused summarization and explanation extraction.

Abstract

We introduce Biased TextRank, a graph-based content extraction method inspired by the popular TextRank algorithm that ranks text spans according to their importance for language processing tasks and according to their relevance to an input "focus." Biased TextRank enables focused content extraction for text by modifying the random restarts in the execution of TextRank. The random restart probabilities are assigned based on the relevance of the graph nodes to the focus of the task. We present two applications of Biased TextRank: focused summarization and explanation extraction, and show that our algorithm leads to improved performance on two different datasets by significant ROUGE-N score margins. Much like its predecessor, Biased TextRank is unsupervised, easy to implement and orders of magnitude faster and lighter than current state-of-the-art Natural Language Processing methods for similar tasks.