Focused Attention Improves Document-Grounded Generation
This addresses the problem of generating accurate and relevant text from documents for applications like content updates and conversational AI, representing a strong specific gain rather than a foundational advance.
The paper tackled document-grounded text generation for Wikipedia updates and dialogue responses by introducing novel adaptations of pre-trained encoder-decoder models with focused attention mechanisms, resulting in at least a 48% increase in BLEU-4 scores over existing methods.
Document grounded generation is the task of using the information provided in a document to improve text generation. This work focuses on two different document grounded generation tasks: Wikipedia Update Generation task and Dialogue response generation. Our work introduces two novel adaptations of large scale pre-trained encoder-decoder models focusing on building context driven representation of the document and enabling specific attention to the information in the document. Additionally, we provide a stronger BART baseline for these tasks. Our proposed techniques outperform existing methods on both automated (at least 48% increase in BLEU-4 points) and human evaluation for closeness to reference and relevance to the document. Furthermore, we perform comprehensive manual inspection of the generated output and categorize errors to provide insights into future directions in modeling these tasks.