Be Consistent! Improving Procedural Text Comprehension using Label Consistency
arXiv:1906.08942
Abstract
Our goal is procedural text comprehension, namely tracking how the properties of entities (e.g., their location) change with time given a procedural text (e.g., a paragraph about photosynthesis, a recipe). This task is challenging as the world is changing throughout the text, and despite recent advances, current systems still struggle with this task. Our approach is to leverage the fact that, for many procedural texts, multiple independent descriptions are readily available, and that predictions from them should be consistent (label consistency). We present a new learning framework that leverages label consistency during training, allowing consistency bias to be built into the model. Evaluation on a standard benchmark dataset for procedural text, ProPara (Dalvi et al., 2018), shows that our approach significantly improves prediction performance (F1) over prior state-of-the-art systems.
NAACL 2019
References in corpus (7)
- Tracking the World State with Recurrent Entity Networks
- A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task
- Learning by Association - A versatile semi-supervised training method for neural networks
- Building Dynamic Knowledge Graphs from Text using Machine Reading Comprehension
- Text2Shape: Generating Shapes from Natural Language by Learning Joint Embeddings
- Tracking State Changes in Procedural Text: A Challenge Dataset and Models for Process Paragraph Comprehension
- Reasoning about Actions and State Changes by Injecting Commonsense Knowledge