SWING: Balancing Coverage and Faithfulness for Dialogue Summarization
This work addresses the challenge of generating accurate and complete summaries for dialogues, which is important for applications like meeting notes or customer service logs, but it is incremental as it builds on existing methods with specific improvements.
The paper tackles the problem of missing information and factual inconsistencies in dialogue summarization by using natural language inference models to improve coverage and faithfulness, achieving effectiveness validated on DialogSum and SAMSum datasets with automatic metrics and human evaluations.
Missing information is a common issue of dialogue summarization where some information in the reference summaries is not covered in the generated summaries. To address this issue, we propose to utilize natural language inference (NLI) models to improve coverage while avoiding introducing factual inconsistencies. Specifically, we use NLI to compute fine-grained training signals to encourage the model to generate content in the reference summaries that have not been covered, as well as to distinguish between factually consistent and inconsistent generated sentences. Experiments on the DialogSum and SAMSum datasets confirm the effectiveness of the proposed approach in balancing coverage and faithfulness, validated with automatic metrics and human evaluations. Additionally, we compute the correlation between commonly used automatic metrics with human judgments in terms of three different dimensions regarding coverage and factual consistency to provide insight into the most suitable metric for evaluating dialogue summaries.