Deep Deception: The story of the spycop network, by the women who uncovered the shocking truth

Product Information

Vrij A. Detecting Lies and Deceit: Pitfalls and Opportunities. 2nd ed. Chichester: John Wiley & Sons, Ltd; 2008. It also followed revelations that undercover officers regularly deceived women into long-term sexual relationships. At least four of the undercover officers are known or alleged to have fathered children with women they met during their deployments. Superior lie-catchers seem to acquire their ability from a personal desire to perform better on their job, no matter what it is [ 5]. It is like any other professional skill or talent, improved through effort, dedication, personal interest, technical knowledge, and training. Thus, such highly skilled lie-catchers result from intense dedication, which is a motivating factor for further research on deception detection. It is reasonable to believe that those levels of accuracy can be approximated or even replicated by a Machine Learning classifier given the correct cues are processed and interpreted.

Burgoon JK, Guerrero LK, Floyd K. Nonverbal communication. 2nd ed. Nonverbal Communication. New York, NY: Routledgr; 2016. 1–509 p. Data gathered from the reviewed documents allows us to safely claim that there has been an increasing interest on deception detection with Machine Learning in the chosen period. In addition, statistical analysis discloses that the approach complexity also increased (see section 3.2 in S6 File) since different modalities were combined and explored to achieve higher performance levels in different scenarios and under various constraints. To the best of our knowledge, this is the first literature review that scrutinizes the application of Machine Learning for deception detection. Trends, gaps, difficulties, results, and opportunities are highlighted to stimulate further studies and new developments in the area.I met with Helen and over the following few months she introduced me to seven other women who had similarly been deceived into long term sexual relationships with undercover police officers. It was now clear that the rogue officer story was far from unique and indeed appeared to reflect a pattern. Alternatively, unsupervised or self-supervised learning happens when the model training does not require labeled data [ 20].

This literature review aims to answer those questions and give a comprehensive overview of the application of Machine Learning for deception detection. We intend to report what researchers have exploited as techniques and approaches, their difficulties, what kind of data they have consumed, and what performance levels they have achieved. Singh A, Thakur N, Sharma A. A review of supervised machine learning algorithms. Proc 10th INDIACom; 2016 3rd Int Conf Comput Sustain Glob Dev INDIACom 2016. 2016;1310–5. Classification tasks rely on algorithms that assign a given class (or label) to a specific data example. Those classes are a limited number of categorical values [ 19]. So, they are not continuous values (while features can be).Authors exploited the Machine Learning arsenal by experimenting with 26 different algorithms. The top five (see section 5.1 in S6 File) is composed by Neural Networks (34 times, 30.09%), Support Vector Machines (SVM) (28 times, 24.78%), Random Forest (20 times, 17.70%), Decision Tree (21 times, 18.58%), and K-Nearest Neighbor (KNN) (10 times, 8.85%). Such algorithms are prevalent and a preference for them is not surprising. AdaBoost (6 papers, 14.63%), Naïve Bayes (6 papers, 14.63%), Logistic Regression (6 papers, 14.63%), and Sequential Minimal Optimization (SMO) (3 papers, 7.32%) come next as the second group of preference. Each of the other 16 (39.09%) algorithms had been exploited once (see section 5 in S6 File). As our contribution to the field, we present a discussion that unfolds in several themes (or dimensions) we consider suitable. Those themes were not chosen. Rather, they arise from the selected documents and represent a general summary of all the efforts analyzed. Those themes are findings themselves. They outline the main topics present in the selected studies regarding the theoretical foundations of deception detection. Authors attempted these approaches to answer to the deception detection problem.