Keeping Notes: Conditional Natural Language Generation with a Scratchpad Mechanism

Abstract

We introduce the Scratchpad Mechanism, a novel addition to the sequence-to-sequence (seq2seq) neural network architecture and demonstrate its effectiveness in improving the overall fluency of seq2seq models for natural language generation tasks. By enabling the decoder at each time step to write to all of the encoder output layers, Scratchpad can employ the encoder as a "scratchpad" memory to keep track of what has been generated so far and thereby guide future generation. We evaluate Scratchpad in the context of three well-studied natural language generation tasks --- Machine Translation, Question Generation, and Text Summarization --- and obtain state-of-the-art or comparable performance on standard datasets for each task. Qualitative assessments in the form of human judgements (question generation), attention visualization (MT), and sample output (summarization) provide further evidence of the ability of Scratchpad to generate fluent and expressive output.

Accepted to ACL 2019

References in corpus (13)

• Sequence to Sequence Learning with Neural Networks
• Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
• Teaching Machines to Read and Comprehend
• A Deep Reinforced Model for Abstractive Summarization
• Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning
• SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive Summarization of Documents
• Layer Normalization
• Incorporating Copying Mechanism in Sequence-to-Sequence Learning
• Question Answering and Question Generation as Dual Tasks
• Language to Logical Form with Neural Attention
• Knowledge Questions from Knowledge Graphs
• A Unified Query-based Generative Model for Question Generation and Question Answering
• Neural Text Generation: A Practical Guide