MReD: A Meta-Review Dataset for Structure-Controllable Text Generation
This provides a domain-specific dataset for structure-controllable text generation, addressing a bottleneck in meta-review analysis, but it is incremental as it builds on existing controllable generation methods.
The paper tackles the problem of limited controllable aspects in text generation by introducing MReD, a dataset of 7,089 meta-reviews with 45k sentences manually annotated into 9 categories, and demonstrates its value through experiments on summarization models.
When directly using existing text generation datasets for controllable generation, we are facing the problem of not having the domain knowledge and thus the aspects that could be controlled are limited. A typical example is when using CNN/Daily Mail dataset for controllable text summarization, there is no guided information on the emphasis of summary sentences. A more useful text generator should leverage both the input text and the control signal to guide the generation, which can only be built with a deep understanding of the domain knowledge. Motivated by this vision, our paper introduces a new text generation dataset, named MReD. Our new dataset consists of 7,089 meta-reviews and all its 45k meta-review sentences are manually annotated with one of the 9 carefully defined categories, including abstract, strength, decision, etc. We present experimental results on start-of-the-art summarization models, and propose methods for structure-controlled generation with both extractive and abstractive models using our annotated data. By exploring various settings and analyzing the model behavior with respect to the control signal, we demonstrate the challenges of our proposed task and the values of our dataset MReD. Meanwhile, MReD also allows us to have a better understanding of the meta-review domain.