Open AccessFFCI: A Framework for Interpretable Automatic Evaluation of Summarization
Author(s) -
Fajri Koto,
Jey Han Lau,
Timothy Baldwin
Publication year - 2022
Publication title -
journal of artificial intelligence research/the journal of artificial intelligence research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.79
H-Index - 123
eISSN - 1943-5037
pISSN - 1076-9757
DOI - 10.1613/jair.1.13167
Subject(s) - automatic summarization , computer science , consistency (knowledge bases) , natural language processing , coherence (philosophical gambling strategy) , precision and recall , sentence , construct (python library) , artificial intelligence , focus (optics) , information retrieval , fluency , similarity (geometry) , range (aeronautics) , semantic similarity , recall , linguistics , mathematics , statistics , physics , philosophy , materials science , composite material , optics , image (mathematics) , programming language
In this paper, we propose FFCI, a framework for fine-grained summarization evaluation that comprises four elements: faithfulness (degree of factual consistency with the source), focus (precision of summary content relative to the reference), coverage (recall of summary content relative to the reference), and inter-sentential coherence (document fluency between adjacent sentences). We construct a novel dataset for focus, coverage, and inter-sentential coherence, and develop automatic methods for evaluating each of the four dimensions of FFCI based on cross-comparison of evaluation metrics and model-based evaluation methods, including question answering (QA) approaches, semantic textual similarity (STS), next-sentence prediction (NSP), and scores derived from 19 pre-trained language models. We then apply the developed metrics in evaluating a broad range of summarization models across two datasets, with some surprising findings.