An analysis of document graph construction methods for AMR summarization

TL;DR

This paper evaluates different methods for merging AMR graphs in document summarization, introducing a new dataset and a novel merging technique that improves content selection accuracy.

Contribution

It presents a new dataset with human-annotated alignments and introduces a novel node merging method that outperforms previous approaches in AMR-based summarization.

Findings

The new merging method significantly improves content selection performance.

The dataset enables systematic evaluation of merge strategies and content selection.

Prior merge strategies are less effective than the proposed method.

Abstract

Meaning Representation (AMR) is a graph-based semantic representation for sentences, composed of collections of concepts linked by semantic relations. AMR-based approaches have found success in a variety of applications, but a challenge to using it in tasks that require document-level context is that it only represents individual sentences. Prior work in AMR-based summarization has automatically merged the individual sentence graphs into a document graph, but the method of merging and its effects on summary content selection have not been independently evaluated. In this paper, we present a novel dataset consisting of human-annotated alignments between the nodes of paired documents and summaries which may be used to evaluate (1) merge strategies; and (2) the performance of content selection methods over nodes of a merged or unmerged AMR graph. We apply these two forms of evaluation to prior work as well as a new method for node merging and show that our new method has significantly better performance than prior work.