Transparent data mining
for Big and Small Data
This book will focus on new emerging data analytics solutions that offer a greater level of transparency than existing solutions. The vast majority of existing algorithms are opaque – that is, the internal algorithmic mechanics are not transparent in that they produce output without making it clear how they have done so. As algorithms increasingly support different aspects of our life, a greater level of transparency is badly needed, not least because discrimination and biases have to be avoided.
Exposing the algorithms software is a challenging task. Although we barely notice them, they are behind a large part of the information we use every day, so rendering algorithms more transparent should improve their usability in various application domains. Thus, transparent solutions are needed to produce more credible and reliable information and services, playing a key role in proactive user engagement by making the results of the data analytical process and its models widely accessible.
The book will include chapters that discuss/address transparent data algorithms for Big and Small Data from different research directions, such as data mining, machine learning, digital ethics, and applied statistics. Book chapters will be written by experts in the emerging field on transparent data algorithms and they will be peer reviewed.
- Introduces an emerging area of data algorithms to academics and practitioners,
- Features real-life use cases in which transparent approaches are essential,
- Provides both the technical and political background to either further the literature in the field or put existing approaches to practical use.
This book aims to provide
- an overview of an emerging area of data analytics algorithms that have profound societal consequences
- specific state-of-the-art solutions
- new application areas that transparent algorithms enable
- an overview of the social and political impact of transparent data analytics solutions
This book will address, but is not limited to, the following topics:
- Transparent data mining solutions with desirable properties (e.g., effective, fully automatic, scalable),
- Experimental findings of transparent solutions tailored to different domain experts,
- Experimental metrics for evaluating algorithmic transparency,
- Societal effects of black box vs. transparent approaches,
- Real-life use cases for black box vs. transparent approaches.
- Abstract submission: May 20th, 2016
- Book Chapter submission: September 9th, 2016
- Review Notification: October 14th, 2016 November 20, 2016
- Camera ready: November 30th , 2016
Each book chapter must comply with the Big Data series formatting guidelines (instructions for authors are available at https://www.springer.com/de/authors-editors/book-authors-editors/manuscript-preparation/5636).
To easily manage submission and reviewing processes we use the easychair system. Book contribution should be submitted in PDF format using the online submission system https://easychair.org/conferences/?conf=glassboxdm2016
This book will come out as part the Studies in Big Data series, published by Springer, which is one of the most prestigious and well regarded.
Tania Cerquitelli (Politecnico di Torino)
Daniele Quercia (Bell Laboratories)
Frank Pasquale (University of Maryland Carey School of Law)
If you have any question or information request on this workshobook, please send an email to tania.cerquitelli AT polito DOT it.