An analysis of document graph construction methods for AMR summarization
This work addresses a bottleneck in AMR summarization for NLP researchers by providing a dataset and evaluation framework, but it is incremental as it builds on existing methods.
The paper tackled the problem of evaluating document graph construction methods for AMR-based summarization by creating a novel dataset with human-annotated alignments, and showed that their new node merging method significantly outperforms prior work.
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.