Covert lie detection using keyboard dynamics

Monaro, Merylin; Galante, Chiara; Spolaor, Riccardo; Li, Qian Qian; Gamberini, Luciano; Conti, Mauro; Sartori, Giuseppe

doi:10.1038/s41598-018-20462-6

Cited by 38 publications

(33 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In order to avoid model hacking, one strategy is to verify that classification accuracy is not changing much among different classes of classifiers (see Monaro et al, 2018) as follows: if similar results are obtained by ML models relying on radically different assumptions, we may be relatively confident that the results are not dependent on such assumptions. Additionally, model stability may be addressed by combining different classifiers into an ensemble classifier that indeed reduces the variance in out-of-sample predictions and therefore gives more reliable predictions.…”

Section: Model-hacking In Machine Learningmentioning

confidence: 99%

“…The above reported examples refer to the recent applications of ML and Deep Learning methods in psychological science that are emerging mainly outside the academic arena. However, the number of experiments reported in academic journals that use ML as analytical tools to complement statistical analysis is also increasing (Kosinski et al, 2013;Monaro et al, 2018;Pace et al, 2019). Machine learning has been successfully applied, for example, in the analysis of imaging data in order to classify psychiatric disorders (Orrù et al, 2012;Vieira et al, 2017), in genetics (Libbrecht and Noble, 2015;Navarin and Costa, 2017), in clinical medicine (Obermeyer and Emanuel, 2016), in forensic sciences (Pace et al, 2019) etc.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning in Psychometrics and Psychological Research

et al. 2020

Self Cite

View full text Add to dashboard Cite

Recent controversies about the level of replicability of behavioral research analyzed using statistical inference have cast interest in developing more efficient techniques for analyzing the results of psychological experiments. Here we claim that complementing the analytical workflow of psychological experiments with Machine Learning-based analysis will both maximize accuracy and minimize replicability issues. As compared to statistical inference, ML analysis of experimental data is model agnostic and primarily focused on prediction rather than inference. We also highlight some potential pitfalls resulting from adoption of Machine Learning based experiment analysis. If not properly used it can lead to over-optimistic accuracy estimates similarly observed using statistical inference. Remedies to such pitfalls are also presented such and building model based on cross validation and the use of ensemble models. ML models are typically regarded as black boxes and we will discuss strategies aimed at rendering more transparent the predictions.

show abstract

Section: Model-hacking In Machine Learningmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Machine Learning in Psychometrics and Psychological Research

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…Cognitive-based lie detection capitalizes on the additional cognitive effort that lying requires compared with truth telling. To date, three main cognitive-based lie techniques have been applied to faked identity detection: the Concealed Information Test (CIT-RT; Verschuere and Kleinberg, 2016 ), the autobiographical Implicit Association Test (aIAT; Agosta and Sartori, 2013 ), and the technique of unexpected questions merged with mouse dynamics or keystroke dynamics (Monaro et al, 2017b , 2018 ). Both the CIT-RT (Verschuere et al, 2011 ) and aIAT (Sartori et al, 2008 ) are memory detection techniques.…”

Section: Introductionmentioning

confidence: 99%

False Identity Detection Using Complex Sentences

et al. 2018

Self Cite

View full text Add to dashboard Cite

The use of faked identities is a current issue for both physical and online security. In this paper, we test the differences between subjects who report their true identity and the ones who give fake identity responding to control, simple, and complex questions. Asking complex questions is a new procedure for increasing liars' cognitive load, which is presented in this paper for the first time. The experiment consisted in an identity verification task, during which response time and errors were collected. Twenty participants were instructed to lie about their identity, whereas the other 20 were asked to respond truthfully. Different machine learning (ML) models were trained, reaching an accuracy level around 90–95% in distinguishing liars from truth tellers based on error rate and response time. Then, to evaluate the generalization and replicability of these models, a new sample of 10 participants were tested and classified, obtaining an accuracy between 80 and 90%. In short, results indicate that liars may be efficiently distinguished from truth tellers on the basis of their response times and errors to complex questions, with an adequate generalization accuracy of the classification models.

show abstract

“…In the most recent literature about deception, a certain number of studies have focused on the detection of faked identity (Monaro et al, 2017c;Monaro et al, 2017d;Monaro et al, 2018a;Verschuere & Kleinberg, 2016). This trend follows the growing need to detect people who cross international borders using fraudulent documents (Hickey, 2015).…”

Section: Introductionmentioning

confidence: 99%

Spotting Faked Identities via Mouse Dynamics Using Complex Questions

Monaro

Gamberini

Giuseppe

2018

Electronic Workshops in Computing

Self Cite

View full text Add to dashboard Cite

The increment of criminals, including terrorists, crossing international borders using faked identities is a crucial issue. This paper validates a computerized technique to spot people who declare false identity information. Forty participants were asked to answer complex questions about their identity, clicking with the mouse on the correct alternative response on the computer screen. Half of the participants answered truthfully, while the others were instructed to lie. As long as the subject responded to questions, mouse dynamics were recorded. Because lying is cognitively demanding, liars had fewer cognitive resources available to analyse complex questions and to compute the response. As result, they showed a bad performance in the task compared with truth-tellers, revealing a greater number of errors, slower reaction times and larger mouse trajectories. Different machine learning classifiers were trained by a cross validation procedure, achieving a classification accuracy up to 90% in detecting liars.

show abstract

Covert lie detection using keyboard dynamics

Cited by 38 publications

References 31 publications

Machine Learning in Psychometrics and Psychological Research

Machine Learning in Psychometrics and Psychological Research

False Identity Detection Using Complex Sentences

Spotting Faked Identities via Mouse Dynamics Using Complex Questions

Contact Info

Product

Resources

About