Justification Narratives for Individual Classifications

By Or Biran, K. McKeown, Biran McKeown. 2014

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Grounding Visual Explanations (2018, background): “Like [2], we aim to generate rationalizations explaining the evidence for a decision as opposed to introspective explanations which aim to explain the intermediate activations of neural networks.”
Attentive Explanations: Justifying Decisions and Pointing to the Evidence (2016, background): “We argue that both are useful; justiﬁcations can provide helpful information for humans in an easily digestible format, whereas introspective explanations can provide insight into a model’s decision process, though it may be harder for a human unfamiliar with deep learning to understand.”
Explainability fact sheets: a framework for systematic assessment of explainable approaches (2019, background): “, a rare feature value) to point the user’s attention in an interesting direction [3, 28, 34] (recallU7where anomalies prompt the user to request an explanation).”
Explainable Artificial Intelligence: a Systematic Review (2020, methodology): “The Justification Narratives method for explainability [184] consists of a simple model-agnostic mapping of the essential values underlying a classification (identified with any feature selection method)”
Generating Visual Explanations (2016, background): “We concentrate on justification explanation systems because they may be more useful to non-experts who do not have knowledge of modern computer vision systems [4].”