A comparison of instance-level counterfactual explanation algorithms for behavioral and textual data: SEDC, LIME-C and SHAP-C

We study the interpretability of predictive systems that use high-dimensonal behavioral and textual data. Examples include predicting product interest based on online browsing data and detecting spam emails or objectionable web content. Recently, counterfactual explanations have been proposed for generating insight into model predictions, which focus on what is relevant to a particular instance. Conducting a complete search to compute counterfactuals is very time-consuming because of the huge dimensionality. To our knowledge, for behavioral and text data, only one model-agnostic heuristic algorithm (SEDC) for finding counterfactual explanations has been proposed in the literature. However, there may be better algorithms for finding counterfactuals quickly. This study aligns the recently proposed Linear Interpretable Model-agnostic Explainer (LIME) and Shapley Additive Explanations (SHAP) with the notion of counterfactual explanations, and empirically benchmarks their effectiveness and efficiency against SEDC using a collection of 13 data sets. Results show that LIME-Counterfactual (LIME-C) and SHAP-Counterfactual (SHAP-C) have low and stable computation times, but mostly, they are less efficient than SEDC. However, for certain instances on certain data sets, SEDC's run time is comparably large. With regard to effectiveness, LIME-C and SHAP-C find reasonable, if not always optimal, counterfactual explanations. SHAP-C, however, seems to have difficulties with highly unbalanced data. Because of its good overall performance, LIME-C seems to be a favorable alternative to SEDC, which failed for some nonlinear models to find counterfactuals because of the particular heuristic search algorithm it uses. A main upshot of this paper is that there is a good deal of room for further research. For example, we propose algorithmic adjustments that are direct upshots of the paper's findings.

[1]  Foster J. Provost,et al.  Enhancing Transparency and Control When Drawing Data-Driven Inferences About Individuals , 2016, Big Data.

[2]  Foster J. Provost,et al.  Explaining Classification Models Built on High-Dimensional Sparse Data , 2016, ArXiv.

[3]  Chandan Singh,et al.  Definitions, methods, and applications in interpretable machine learning , 2019, Proceedings of the National Academy of Sciences.

[4]  Foster J. Provost,et al.  Explaining Data-Driven Document Classifications , 2013, MIS Q..

[5]  Ping Li,et al.  One Permutation Hashing for Efficient Search and Learning , 2012, ArXiv.

[6]  Tim Miller,et al.  Explanation in Artificial Intelligence: Insights from the Social Sciences , 2017, Artif. Intell..

[7]  Marcel van Gerven,et al.  Explanation Methods in Deep Learning: Users, Values, Concerns and Challenges , 2018, ArXiv.

[8]  F. Maxwell Harper,et al.  The MovieLens Datasets: History and Context , 2016, TIIS.

[9]  Foster J. Provost,et al.  Mining Massive Fine-Grained Behavior Data to Improve Predictive Analytics , 2016, MIS Q..

[10]  Dong Nguyen,et al.  Comparing Automatic and Human Evaluation of Local Explanations for Text Classification , 2018, NAACL.

[11]  G. A. Miller.  THE PSYCHOLOGICAL REVIEW THE MAGICAL NUMBER SEVEN, PLUS OR MINUS TWO: SOME LIMITS ON OUR CAPACITY FOR PROCESSING INFORMATION 1 , 1956 .

[12]  Janis Klaise,et al.  Interpretable Counterfactual Explanations Guided by Prototypes , 2019, ECML/PKDD.

[13]  Chun-Nan Hsu,et al.  Mining Skewed and Sparse Transaction Data for Personalized Shopping Recommendation , 2004, Machine Learning.

[14]  T. Graepel,et al.  Private traits and attributes are predictable from digital records of human behavior , 2013, Proceedings of the National Academy of Sciences.

[15]  Foster J. Provost,et al.  Predictive Modeling With Big Data: Is Bigger Really Better? , 2013, Big Data.

[16]  Izak Benbasat,et al.  Explanations From Intelligent Systems: Theoretical Foundations and Implications for Practice , 1999, MIS Q..

[17]  Morten Wang Fagerland,et al.  The McNemar test for binary matched-pairs data: mid-p and asymptotic are better than exact conditional , 2013, BMC Medical Research Methodology.

[18]  Vaclav Petricek,et al.  Recommender System for Online Dating Service , 2007, ArXiv.

[19]  Susan T. Dumais,et al.  Inductive learning algorithms and representations for text categorization , 1998, CIKM '98.

[20]  Peter A. Flach,et al.  Counterfactual Explanations of Machine Learning Predictions: Opportunities and Challenges for AI Safety , 2019, SafeAI@AAAI.

[21]  Galit Shmueli,et al.  Analyzing Behavioral Big Data: Methodological, practical, ethical, and moral issues , 2016 .

[22]  Ethan L. Schreiber.  Optimal Multi-Way Number Partitioning , 2018, J. ACM.

[23]  Been Kim,et al.  Towards A Rigorous Science of Interpretable Machine Learning , 2017, 1702.08608.

[24]  Theodoros Evgeniou,et al.  A benchmarking study of classification techniques for behavioral data , 2019, International Journal of Data Science and Analytics.

[25]  Foster Provost,et al.  Dimensionality Reduction Via Matrix Factorization for Predictive Modeling from Large, Sparse Behavioral Data , 2015 .

[26]  Chris Russell,et al.  Efficient Search for Diverse Coherent Explanations , 2019, FAT.

[27]  Scott Lundberg,et al.  A Unified Approach to Interpreting Model Predictions , 2017, NIPS.

[28]  M. Kosinski,et al.  Psychological targeting as an effective approach to digital mass persuasion , 2017, Proceedings of the National Academy of Sciences.

[29]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[30]  David Martens,et al.  DEPARTMENT OF ENGINEERING MANAGEMENT Classification over bipartite graphs through projection , 2015 .

[31]  Daniele Quercia,et al.  Our Twitter Profiles, Our Selves: Predicting Personality with Twitter , 2011, 2011 IEEE Third Int'l Conference on Privacy, Security, Risk and Trust and 2011 IEEE Third Int'l Conference on Social Computing.

[32]  Franco Turini,et al.  Local Rule-Based Explanations of Black Box Decision Systems , 2018, ArXiv.

[33]  Alexander Binder,et al.  Evaluating the Visualization of What a Deep Neural Network Has Learned , 2015, IEEE Transactions on Neural Networks and Learning Systems.

[34]  Alex Alves Freitas,et al.  Comprehensible classification models: a position paper , 2014, SKDD.

[35]  Chandan Singh,et al.  Definitions, methods, and applications in interpretable machine learning , 2019, Proceedings of the National Academy of Sciences.

[36]  Klaus-Robert Müller,et al.  "What is relevant in a text document?": An interpretable machine learning approach , 2016, PloS one.

[37]  Thorsten Joachims,et al.  Text Categorization with Support Vector Machines: Learning with Many Relevant Features , 1998, ECML.

[38]  Carlos Guestrin,et al.  "Why Should I Trust You?": Explaining the Predictions of Any Classifier , 2016, ArXiv.

[39]  Bart Baesens,et al.  Comprehensible Credit Scoring Models Using Rule Extraction from Support Vector Machines , 2007, Eur. J. Oper. Res..

[40]  Trevor Darrell,et al.  Grounding Visual Explanations , 2018, ECCV.

[41]  Chris Russell,et al.  Counterfactual Explanations Without Opening the Black Box: Automated Decisions and the GDPR , 2017, ArXiv.

[42]  Krishna P. Gummadi,et al.  Measurement and analysis of online social networks , 2007, IMC '07.

[43]  David Martens,et al.  Who cares about your Facebook friends? Credit scoring for microfinance , 2015 .

[44]  M. Kosinski,et al.  Musical Preferences Predict Personality: Evidence From Active Listening and Facebook Likes , 2018, Psychological science.