Transparent Data Mining for Big and Small Data

1. Front Matter

   Pages i-xv

   PDF

2. Transparent Mining

   1. Front Matter

      Pages 1-1

      PDF

   2. The Tyranny of Data? The Bright and Dark Sides of Data-Driven Decision-Making for Social Good

      • Bruno Lepri, Jacopo Staiano, David Sangokoya, Emmanuel Letouzé, Nuria Oliver

      Pages 3-24

   3. Enabling Accountability of Algorithmic Media: Transparency as a Constructive and Critical Lens

      • Nicholas Diakopoulos

      Pages 25-43

   4. The Princeton Web Transparency and Accountability Project

      • Arvind Narayanan, Dillon Reisman

      Pages 45-67

3. Algorithmic Solutions

   1. Front Matter

      Pages 69-69

      PDF

   2. Algorithmic Transparency via Quantitative Input Influence

      • Anupam Datta, Shayak Sen, Yair Zick

      Pages 71-94

   3. Learning Interpretable Classification Rules with Boolean Compressed Sensing

      • Dmitry M. Malioutov, Kush R. Varshney, Amin Emad, Sanjeeb Dash

      Pages 95-121

   4. Visualizations of Deep Neural Networks in Computer Vision: A Survey

      • Christin Seifert, Aisha Aamir, Aparna Balagopalan, Dhruv Jain, Abhinav Sharma, Sebastian Grottel et al.

      Pages 123-144

4. Regulatory Solutions

   1. Front Matter

      Pages 145-145

      PDF

   2. Beyond the EULA: Improving Consent for Data Mining

      • Luke Hutton, Tristan Henderson

      Pages 147-167

   3. Regulating Algorithms’ Regulation? First Ethico-Legal Principles, Problems, and Opportunities of Algorithms

      • Giovanni Comandè

      Pages 169-206

   4. AlgorithmWatch: What Role Can a Watchdog Organization Play in Ensuring Algorithmic Accountability?

      • Matthias Spielkamp

      Pages 207-215

5. Erratum

   • Tania Cerquitelli, Daniele Quercia, Frank Pasquale

   Pages E1-E1

This book focuses on new and emerging data mining solutions that offer a greater level of transparency than existing solutions. Transparent data mining solutions with desirable properties (e.g. effective, fully automatic, scalable) are covered in the book. Experimental findings of transparent solutions are tailored to different domain experts, and experimental metrics for evaluating algorithmic transparency are presented. The book also discusses societal effects of black box vs. transparent approaches to data mining, as well as real-world use cases for these approaches.
As algorithms increasingly support different aspects of modern life, a greater level of transparency is sorely needed, not least because discrimination and biases have to be avoided. With contributions from domain experts, this book provides an overview of an emerging area of data mining that has profound societal consequences, and provides the technical background to for readers to contribute to the field or to put existing approaches to practical use.

• Transparent Predictive Models
• Glass-box Algorithms
• Black-box Algorithms
• Transparent vs Opaque Algorithms
• Automated Decision Making
• Big Data Paradigm Shift
• algorithm analysis and problem complexity
• complexity

• Department of Control and Computer Engineering, Politecnico di Torino, Torino, Italy

  Tania Cerquitelli

• Bell Laboratories, Cambridge, United Kingdom

  Daniele Quercia

• Carey School of Law, University of Maryland, Baltimore, USA

  Frank Pasquale

Policies and ethics