Explaining Data-Driven Document Classifications

By David Martens, Foster J. Provost. 2014

319 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Management Science and Operations Research, Safety, Robustness & Interpretability, Information Extraction

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

In Management Science and Operations Research

In Safety, Robustness & Interpretability

In Information Extraction

Citations in the corpus

Explanation in Human-AI Systems: A Literature Meta-Review, Synopsis of Key Ideas and Publications, and Bibliography for Explainable AI (2019, neutral): “…and Mooney, 2005; Biran and McKeown, 2014, 2017; Craven and Shavlik, 1996; Lakkaraju, Kamar, Caruana, and Leskovec, 2017; Lou, Caruana, and Gehrke, 2012; Martens, Huysmans, Setiono, Vanthienen, and Baesens, 2008; Martens and Provost, 2014; Montavon, Lapuschkin, Binder, Samek, and Müller-(2017).”
Explaining Classification Models Built on High-Dimensional Sparse Data (2016, methodology): “Prior work suggests that when users do not understand the workings of a classiﬁcation model, they can be reluctant to use it, even if the model is known to improve decision performance (Kayande et al., 2009; Martens & Provost, 2014).”
Enhancing Transparency and Control When Drawing Data-Driven Inferences About Individuals (2016, background): “These massive models are used to drive decisions for millions of individuals; they can learn trends that are incorrect, or perpetuate social biases, and result in social and emotional outcomes that are harmful to people and social groups.”
ViCE: visual counterfactual explanations for machine learning models (2020, methodology): “In order to extend the algorithm proposed in [12], the entire dataset is discretized by fitting a Gaussian on each of the features and splitting the values into n bins such that themiddle n-2 capture”
Predictive Multiplicity in Classification (2019, background): “Credit Scoring : Say we must explain the prediction of a credit scoring model to a person [by producing e.g., a counterfactual explanation 31, 50].”
Strongly agree or strongly disagree?: Rating features in Support Vector Machines (2016, methodology): “…popular approach has been to extract easy to understand structures from powerful yet black box type classi ers, such as if then rules, decision trees and decision tables (Baesens et al. 2003, Bertsimas et al. 2011, Martens et al. 2007, Martens and Provost 2013, Orsenigo and Vercellis 2003, 2004).”
A data-driven approach to predict the success of bank telemarketing (2014, background): “Opening the black-box allows to explaining how the model makes the decisions and improves the acceptance of prediction models by the domain experts, as shown in [20].”
Explaining Explanations in AI (2018, methodology): “ues arise in the fitting of models: if the domain the model is fitted to does not capture examples relevant to the intended audience, then it is unlikely to be useful. The first approach described by Martens and Provost (2013) is explicitly designed for use on discrete data, and focuses on the parti”
Efficient Search for Diverse Coherent Explanations (2019, methodology): “nimal change such that the response of the algorithm changes e.g. “You were denied a loan because you have an income of $30,000, if you had an income of $45,000 you would have been offered the loan.” Martens and Provost (2013) was the first to propose the use of this technique in the context of remo”
A survey of algorithmic recourse: definitions, formulations, solutions, and prospects (2020, neutral): “03) SEDC [119] E heuristic query (2015.”
A Workflow for Visual Diagnostics of Binary Classifiers using Instance-Level Explanations (2017, background): “Martens and Provost [24] provide an algorithm for computing minimal explanations for binary input data.”
Explainable Artificial Intelligence: Objectives, Stakeholders, and Future Research Opportunities (2020, background): “Besides only very few examples (e.g. Eiras-Franco et al., 2019; Giboney et al., 2015; Martens & Provost, 2014), since then most of the publications 1 on explain-ability of AI systems, or “Explainable Artificial Intelligence” (XAI), have been published outside of the information systems community,…”