Facet-Aware Evaluation for Extractive Summarization

TL;DR

This paper introduces a facet-aware evaluation method for extractive summarization that assesses information coverage more effectively than traditional lexical overlap metrics, using a new dataset and demonstrating better correlation with human judgment.

Contribution

It proposes a novel facet-aware evaluation framework, constructs an extractive dataset from CNN/Daily Mail, and shows improved evaluation accuracy over ROUGE.

Findings

Facet-aware evaluation correlates better with human judgment.

Enables fine-grained and comparative analysis of summarization methods.

Reveals insights into the performance of state-of-the-art models.

Abstract

Commonly adopted metrics for extractive summarization focus on lexical overlap at the token level. In this paper, we present a facet-aware evaluation setup for better assessment of the information coverage in extracted summaries. Specifically, we treat each sentence in the reference summary as a \textit{facet}, identify the sentences in the document that express the semantics of each facet as \textit{support sentences} of the facet, and automatically evaluate extractive summarization methods by comparing the indices of extracted sentences and support sentences of all the facets in the reference summary. To facilitate this new evaluation setup, we construct an extractive version of the CNN/Daily Mail dataset and perform a thorough quantitative investigation, through which we demonstrate that facet-aware evaluation manifests better correlation with human judgment than ROUGE, enables fine-grained evaluation as well as comparative analysis, and reveals valuable insights of state-of-the-art summarization methods. Data can be found at https://github.com/morningmoni/FAR.