How to Evaluate a Summarizer: Study Design and Statistical Analysis for Manual Linguistic Quality Evaluation
This work addresses methodological inconsistencies in summarization evaluation, which is crucial for researchers in NLP to ensure reliable system comparisons.
The paper tackles the lack of standardization in manual evaluation for text summarization by conducting a survey and experiments, revealing that evaluation methods vary by aspect and that current practices inflate type I error rates up to eight-fold.
Manual evaluation is essential to judge progress on automatic text summarization. However, we conduct a survey on recent summarization system papers that reveals little agreement on how to perform such evaluation studies. We conduct two evaluation experiments on two aspects of summaries' linguistic quality (coherence and repetitiveness) to compare Likert-type and ranking annotations and show that best choice of evaluation method can vary from one aspect to another. In our survey, we also find that study parameters such as the overall number of annotators and distribution of annotators to annotation items are often not fully reported and that subsequent statistical analysis ignores grouping factors arising from one annotator judging multiple summaries. Using our evaluation experiments, we show that the total number of annotators can have a strong impact on study power and that current statistical analysis methods can inflate type I error rates up to eight-fold. In addition, we highlight that for the purpose of system comparison the current practice of eliciting multiple judgements per summary leads to less powerful and reliable annotations given a fixed study budget.