Do Human Rationales Improve Machine Explanations?
By Julia Strout, Ye Zhang, Raymond J. Mooney. 2019
64 citations recorded by the source corpus.
Work on “learning with rationales” shows that humans providing explanations to a machine learning system can improve the system’s predictive accuracy. However, this work has not been connected to work in “explainable AI” which concerns machines explaining their reasoning to humans. In this work, we show that learning with rationales can also improve the quality of the machine’s explanations as evaluated by human judges. Specifically, we present experiments showing that, for CNN-based text classification, explanations generated using “supervised attention” are judged superior to explanations generated using normal unsupervised attention.
Subfields: Artificial Intelligence
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Ana Marasovic
- Sarah Wiegreffe
- Yoav Goldberg
- Noah A. Smith
- Mohit Bansal
- Alon Jacovi
- Yejin Choi
- Yulia Tsvetkov
Citations in the corpus
- Interpretations are useful: penalizing explanations to align neural networks with prior knowledge (2019, background): “This work on annotator rationales was recently extended to show improved explanations (not accuracy) in particular types of CNNs (Strout et al., 2019).”
- ERASER: A Benchmark to Evaluate Rationalized NLP Models (2019, background): “Strout et al. (2019) have demonstrated that providing this model with target ra- tionales at train time results in the model providing rationales at test time that are preferred by humans (compared to rationales provided when the model learns to weight sentences in an end-to-end fashion).”
- Towards Faithfully Interpretable NLP Systems: How Should We Define and Evaluate Faithfulness? (2020, background): “ample, consider the case of interpretation via posthoc text generation—where an additional “generator” component outputs a textual explanation of the model’s decision, and the generator is learned with supervision of textual explanations (Zaidan and Eisner, 2008; Rajani et al., 2019; Strout et al., ”
- Teach Me to Explain: A Review of Datasets for Explainable Natural Language Processing (2021, background): “[116] Julia Strout, Ye Zhang, and Raymond Mooney.”
- Learning to Faithfully Rationalize by Construction (2020, background): “Prior work has exploited such signal during training (Zhang et al., 2016; Strout et al., 2019; Small et al., 2011).”