V. S. Subrahmanian (editor) - Detecting Trust and Deception in Group Interaction

The purpose of the Computation and International Security book series is to establish the state of the art and set the course for future research in computational approaches to international security. The scope of this series is broad and aims to look at computational research that addresses topics in counter-terrorism, counter-drug, transnational crime, homeland security, cyber-crime, public policy, international conflict, and stability of nations. Computational research areas that interact with these topics include (but are not restricted to) research in databases, machine learning, data mining, planning, artificial intelligence, operations research, mathematics, network analysis, social networks, computer vision, computer security, biometrics, forecasting, and statistical modeling. The series serves as a central source of reference for information and communications technology that addresses topics related to international security. The series aims to publish thorough and cohesive studies on specific topics in international security that have a computational and/or mathematical theme, as well as works that are larger in scope than survey articles and that will contain more detailed background information. The series also provides a single point of coverage of advanced and timely topics and a forum for topics that may not have reached a level of maturity to warrant a comprehensive textbook.

More information about this series at http://www.springer.com/series/11955

This Springer imprint is published by the registered company Springer Nature Switzerland AG

The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

With the rapid increase in the amount of video available online, there is growing interest in analyzing the content of such videos.

Social scientists have an interest in analyzing videos of a group of people interacting together in order to better understand the factors that lead to a host of behaviors exhibited in group interactions. In particular, social scientists would like to understand facial and vocal cues that are linked to answers to questions such as: Does person X like person Y? Does person X trust person Y? Is person X more dominant than person Y? Is person X being deceptive? The use of carefully curated, data-driven studies such as these are incredibly helpful in validating existing theories linking facial and vocal cues to such behaviors.

At the same time, numerous corporations and governments have an interest in the automated analysis of videos of group interactions. Recent advances in machine learning offer the promise of building end-to-end systems that analyze video and automatically predict the answers to questions such as those listed above. We note that prediction of a behavioral event and understanding the factors that are cues to the occurrence of a behavior are not the same thing.

This edited book brings together a series of chapters that build on a novel dataset created by the authors and their students and collaborators. By focusing on carefully collected video of a popular face-to-face game called The Resistance, we present a comprehensive overview of the power of both social science theory and computational modeling in understanding and predicting behavior during interactions among a group of people.

We are grateful to numerous people who have helped produce this book. First, we would like to express our gratitude to the Army Research Office for funding much of the work reported in this book under Grant W911NF-16-1-0342. In particular, we would like to thank Dr. Purush Iyer, Dr. Edward Palozzolo, and Dr. Lisa Troyer for their very strong support to this work, along with a continuous stream of valuable comments, encouragement, and advice. Thanks are also due to Dr. Addison Bohannon, Dr. Liz Bowman, Dr. Javier Garcia, and Dr. Jean Vettel for sharing a host of interesting and useful thoughts. Of course, we are grateful to the PIs on this project from numerous institutions: Profs. Larry Davis, Jure Leskovec, Miriam Metzger, and Jay Nunamaker. We are also deeply grateful to Chongyang Bai for his hard work in formatting the chapters and helping to pull all the chapters submitted into a single document. Finally, we express our sincere gratitude to many of the students and postdocs who helped shape this effort and generate the results presented in this book.