Deception Detection Research Articles

Hoax news validation using similarity algorithms

News that is presented every day on social media dramatically affects the feelings, feelings, thoughts, or even actions of a person or group. Hoax News is one of them which is disturbing the public and raising noise in various fields, ranging from politics, culture, security, and order, to the economy. Inseparable from social media users. How every day, there is information on social media, which is not necessarily true so that people are provoked by hoax on social media. The news detection system in this study was designed using Unsupported Learning so that it does not require data training. The system was built using the Equation algorithm to calculate the validity of document similarity. Extraction results used to search for content related to user input using a detection engine, then the similarity value and the time needed to utilize hoax news are calculated. System validation testing by using a four text similarity algorithm called the Equation algorithm, the Levenshtein algorithm, the Smith-Waterman algorithm, the Damerau Levenshtein algorithm; this algorithm is used to find the best analytical solution of news hoaxes and submissions needed to find the news hoax password. The final results of the deception detection research using a script that has been done for Validation using an algorithm, get the value of accuracy in detection using the Smith-Waterman algorithm, which produces an accuracy value of text similarity of 99.29% and can be used a process of 6, 57 seconds, followed by the second sequence that is the similarity algorithm produces an accuracy of 75% and requires a processing time of 4.94 seconds, then the third sequence is the Levenshtein algorithm with an accuracy of 55.02% and requires a processing time of 5.49 seconds, and is used today is Damerau Levenshtein algorithm is 55.02% and requires a processing time of 7.54%. The results of research tests on this text can conclude the more text on the detection engine, the higher the verification value and the higher the time needed to process hoax news.

설문 기반 대응방안을 사용한 복합시행 프로토콜의 재평가

Traditional deception detection methods had examined the difference of one’s autonomic physiological responses through asking crime-related and crime-unrelated questions. There has been a continuing controversy regarding the accuracy and validity of the test, and thus, many researchers were motivated to explore and develop alternative efficient methods of detection in which one of them is known as P300-based Complex Trial Protocol (CTP). The P300-based CTP detects deception through comparing the P300 amplitudes between probe and irrelevant stimuli and is known as a counterstrategy of countermeasures. However, many previous studies have used countermeasures created from Rosenfeld et al.’s work (2008).The present study initially conducted a survey asking open-ended questions about the countermeasure use to acquire participant-oriented countermeasures for the main experiment. Then, the study aimed to evaluate whether the CTP can accurately detect deception even in the use of survey-based countermeasures. We firstly selected a set of participant-oriented countermeasures through survey questions. Then, a total of 50 participants were divided into three groups (innocent, guilty, and countermeasures) and performed the CTP. Those assigned to the countermeasures group covertly performed mental countermeasures during the CTP. The results of P300 amplitude analysis revealed that the guilty group’s P300 amplitude of probe stimuli was significantly larger than that of irrelevant stimuli. Countermeasures group also had a significantly larger P300 amplitude for probe stimuli compared to irrelevant stimuli, even in the use of countermeasures. The results of bootstrapped amplitude difference (BAD) showed a detection accuracy rate of 81.25%, 82.35%, 82.35% for the innocent, guilty, and countermeasures groups, respectively. These findings demonstrate that the CTP can obtain a high detection rate in participant-oriented countermeasures and suggest the potential use of the CTP in the

Role of sentiment analysis in social media security and analytics

AbstractSocial media, in recent times, has with eased an explosion of data with so many social media platforms available to interact and express opinions freely. This has led to easy access to the privacy of social media users which raise broader security concerns and issues. The present paper provides an overview of various sentiment analysis approaches and techniques for social media security and analytics. The multiple security application domains like deception detection, anomaly detection, risk management, and disaster relief have been identified where sentiment analysis is used for social media security. An in‐depth study on security issues related to data provenance, distrust, e‐commerce security, consumer security breaches, market surveillance, credibility, and risk assessment in social media have been presented. A comparison of various techniques, methodologies, dataset, and application domain where sentiment analysis is used has been discussed. The present work discusses the results of different machine learning techniques based on the performance metrics that have been used for the implementation of sentiment analysis in the respective security domains. It identifies the various gaps, issues, and the recent advancements in the field and presents a line of work that needs to be carried forward in future.This article is categorized under: Commercial, Legal, and Ethical Issues > Security and Privacy Technologies > Machine Learning Technologies > Prediction

The Role of External Audit in View of Sudden Financial Collapses on the Jordanian Commercial Banks Sector

The study aims to determine the role of external audit in view of sudden financial collapse son Jordanian commercial banks. To achieve this, the researcher designed a questionnaire consists of (40) paragraphs.The study results show that there is a statistically significant impact (the first hypothesis) of external audit on sudden financial collapses on Jordanian commercial banks. The value of the relationship between the two variables is 0.638 which is statistically significant, because calculated f value is equal to (75.31) which is statistically significant as (0.000) value, which is less than (5%).The study also indicates that there is a statistically significant effect (the second hypothesis) of external audit on the compliance of the audit profession rules on Jordanian commercial banks. The value of the relationship between the two variables is 0.615 which is statistically significant, because calculated f value is equal to (60.09) which is statistically significant as (0.000) value, which is less than (5%).The findings suggest that there is a statistically significant effect (the third hypothesis) of external audit on the detection of deception and fraud on Jordanian commercial banks. Where the value of the relationship between the two variables is (0.673) which is statistically significant, because calculated f value is equal to 91.07 which is statistically significant as (0.000) value, which is less than (5%).

An empirical evaluation of interpersonal deception theory in a real-world, high-stakes environment

Purpose The study of deception and the theories, which have been developed have relied heavily on laboratory experiments in controlled environments, using American college students participating in mock scenarios. The purpose of this paper is to validate previous deception detection research in a real-world, high stakes environment where the unit of analysis is the question–response pair. Design/methodology/approach The study used previously confirmed linguistic and paralinguistic speech cues and the constructs of deception in an attempt to validate a leading deception theory, interpersonal deception theory (IDT). A combination of descriptive and predictive analysis was conducted to best understand the relationship between speech cues and changes in the subjects’ behavior. Findings The result validates IDT with mixed results on individual measures and their constructs. However, there is clear evidence across the 711 question-response pairs that not only was it possible to differentiate truth from deceptive behavior but also patterns of behavior can be seen over time. Research limitations/implications Because of the real-world nature of the study, it is difficult to generalize the results to a larger population. However, one implication for future research is the development of methods to capture, process and prepare raw speech into data ready for analysis. Originality/value This paper attempts to fill the gap between the controlled mock scenarios and the harsh reality of real-world deception.

Applying the verifiability approach to deception detection in alibi witness situations

The application of alibi witness scenarios to deception detection has been overlooked. Experiment 1 was a study of the verifiability approach in which truth-telling pairs completed a mission together, whereas in lying pairs one individual completed this mission alone and the other individual committed a mock theft. All pairs were instructed to convince the interviewer that they completed the mission together by writing individual statements on their own followed by a collective statement together as a pair. In the individual statements, truth-telling pairs provided more checkable details that demonstrated they completed the mission together than lying pairs, whereas lying pairs provided more uncheckable details than truth-telling pairs. The collective statements made truth-telling pairs provide significantly more checkable details that demonstrated they were together in comparison to the individual statements, whereas no effect was obtained for lying pairs. Receiver Operating Characteristic curves revealed high accuracy rates for discriminating between truths and lies using the verifiability approach across all statement types. Experiment 2 was a lie detection study whereby observers' abilities to discriminate between truths and lies using the verifiability approach were examined. This revealed that applying the verifiability approach to collective statements improved observers' ability to accurately detect deceit. We suggest that the verifiability approach could be used as a lie detection technique and that law enforcement policies should consider implementing collective interviewing.

THE RELATIONSHIP BETWEEN THE ATTITUDE OF LAW ENFORCEMENT OFFICERS AND CIVILIANS TO LIES WITH THE SUCCESS OF ITS RECOGNITION

The article is devoted to the description of attitude to lies and success of its detection in groups of civil recipients and police. Results of two empirical studies are presented in the article. At the first study (n = 120) was found that civil recipients are better in deception detection if they have epistemological understanding of concept of lie. Those recipients who understand the lie from a moral position, condemn it – are worse in lie detection. Those recipients who considers the main motive of lies is «to create a good effect» are also successful in deception detection. Perhaps this motive is projection, and people who have this motive are more attentive to the communicant, spend more effort to recognize verbal and non-verbal cues. The comparison group of recipients who rely on scientifically confirmed and unconfirmed cues of deception was shown, that scientifically confirmed group is better in lie detection. In a group of civil recipients it was shown that attitudes to different types of lies are correlated. It was also found that the better a person evaluates himself as a successful liar, the more often he uses lie in life and the more positive attitude to lies. However, the attitude to different types of lies is not correlated with the success of detection lie, truth and messages (lie+rtuth) - only with self-esteem as a successful or unsuccessful recipient. In the second study (n = 62) it was found that lie detection in police is not correlated with age, work experience, attitude to white lies. The police, in general, are more aware of the cues of lie, more often civilians rely on scientifically confirmed, in particular, verbal cues. However focus on scientifically confirmed cues of lie is associated with prejudgment in police. The main motive of lies «to create a good effect» named by the police improves lie detection, like in civil group. But understanding lie as ontological or moral position is not associated with lie detection in police. In general, both groups detect lies at the level of random guessing (50%), which is consistent with other studies. Differences in success in detection of messages (lie+rtuth) are due to bias. Biases, probably, are more associated with attitude to lie.

Towards Real-Time Eyeblink Detection in the Wild: Dataset, Theory and Practices

Effective and real-time eyeblink detection is of wide-range applications, such as deception detection, drive fatigue detection, face anti-spoofing. Despite previous efforts, most of existing focus on addressing the eyeblink detection problem under constrained indoor conditions with relative consistent subject and environment setup. Nevertheless, towards practical applications, eyeblink detection in the wild is highly preferred, and of greater challenges. In this paper, we shed the light to this research topic. A labelled eyeblink in the wild dataset (i.e., HUST-LEBW) of 673 eyeblink video samples (i.e., 381 positives, and 292 negatives) is first established. These samples are captured from the unconstrained movies, with the dramatic variation on face attribute, head pose, illumination condition, imaging configuration, etc. Then, we formulate eyeblink detection task as a binary spatial-temporal pattern recognition problem. After locating and tracking human eyes using SeetaFace engine and KCF (Kernelized Correlation Filters) tracker respectively, a modified LSTM model able to capture the multi-scale temporal information is proposed to verify eyeblink. A feature extraction approach that reveals the appearance and motion characteristics simultaneously is also proposed. The experiments on HUST-LEBW reveal the superiority and efficiency of our approach. The comparisons with the existing state-of-the-art methods validate the advantages of our manner for eyeblink detection in the wild.

Design and Implementation of Deception Detection System Based on Reliable Facial Expression

Design and Implementation of Deception Detection System Based on Reliable Facial Expression

JUDGING ANTIGONE IN OLD AGE: (NEURO)PSYCHOLOGY AND LAW IN THE TWENTY-FIRST CENTURY

Although medicine, psychology and law have a long history behind them (Frank, 1930; Kapardis, 2009), and several issues may emerge at the interface of these scientific fields, new specific disciplines have been created to fill the research gap, such as legal psychology and forensic psychology (Taylor, 2019). In this direction, forensic neuropsychology combines not only psychometric testing, but also interviews, behavioral observations, and record reviews as additional sources of information for answering cognitive functioning questions in legal and forensic settings (Fink, 2007). Additionally, the emergence of a new promising interdisciplinary field, neurolaw, explores how neuroscience and advanced techniques can be used in the legal system (Meynen, 2014; van Dongen & Franken, 2019). Although there are still a plethora of topics that are not adequately described and need further research in neuropsychology, several attempts have been made to collect data for different research questions linked to this complex field (Canter & Žukauskiene, 2019). Methods and protocols regarding the processes of criminal profiling, the examination of cognitive biases in legal decision making, interrogations and confessions, detection of deception, eyewitness memory and false memories-distortions, cognition and emotions of child witnesses, and competence to stand trial, are still not clear (Bornstein & Miller, 2016; Brewer & Douglass, 2019).

Why Won’t You Answer the Question? Mass-Mediated Deception Detection After Journalists’ Accusations of Politicians’ Evasion

Abstract Journalists often accuse politicians of dodging questions. Truth-default theory (TDT) predicts that when journalists serve as de facto deception detectors, the audience will process the messaging through a cognitive sequence that lowers the perceived trustworthiness of the politician. Conversely, the public’s perception of the media as being generally hostile and biased in their reporting could make a journalist’s allegation of evasion enhance the politician’s credibility. We constructed political TV interviews in which a journalist falsely accused a politician of evasiveness. Consistent with serial multiple mediation as proposed by TDT, in Study 1 (N = 210 U.S. voters) a journalist’s allegation triggered suspicion, which increased perceived dodging, resulting in voters distrusting the politician. Absent a journalist’s allegation, however, people remained in their truth-default state toward the politician. Study 2 (N = 429) replicated the Study 1 results, and conditional process modeling revealed that the effect was moderated by rumination.

Interview based connectivity analysis of EEG in order to detect deception

Deception is mentioned as an expression or action which hides the truth and deception detection as a concept to uncover the truth. In this research, a connectivity analysis of Electro Encephalography study is presented regarding cognitive processes of an instructed liar/truth-teller about identity during an interview. In this survey, connectivity analysis is applied because it can provide unique information about brain activity patterns of lying and interaction among brain regions. The novelty of this paper lies in applying an open-ended questions interview protocol during EEG recording. We recruited 40 healthy participants to record EEG signal during the interview. For each subject, whole-brain functional and effective connectivity networks such as coherence, generalized partial direct coherence and directed directed transfer function, are constructed for the lie-telling and truth-telling conditions. The classification results demonstrate that lying could be differentiated from truth-telling with an accuracy of 86.25% with the leave-one-person-out method. Results show functional and effective connectivity patterns of lying for the average of all frequency bands are different in regions from that of truth-telling. The current study may shed new light on neural patterns of deception from connectivity analysis view point.

Teaching Students About Sender and Receiver Variability in Lie Detection

Deception detection is a topic covered in many psychology and communication courses. We designed and implemented an engaging class activity to facilitate the students’ learning of several key concepts related to sender and receiver variability in lie detection. The pedagogical effectiveness of the activity was measured. In line with previous research, the students had difficulty acknowledging and correcting their own biases. However, the teaching activity was successful in correcting their general beliefs and facilitating learning.

Incorporating feature representation into BiLSTM for deceptive review detection

Consumers are increasingly influenced by product reviews when purchasing goods or services. At the same time, deceptive reviews usually mislead users. It is inefficient and inaccurate to manually identify deceptive reviews in massive reviews. Therefore, automatically identifying deceptive reviews has become a research trend. Most of existing methods are less effective since they are lack of deeply understanding of reviews. We propose a neural network method with bidirectional long short-term memory (BiLSTM) and feature combination to learn the representation of deceptive reviews. We conduct a large amount of experiments and demonstrate the effectiveness of our proposed method. Specifically, in the mixed-domain detection experiment, the results prove that our model is effective by making comparisons with other neural network-based methods. BiLSTM gives more than 3% improvement in F1 score compared with the most advanced neural network method. Since feature selection plays an important role in this direction, we combine features to improve the performance. Then we get 87.6% F1 value which outperforms the state-of-the-art method. Moreover, in the cross-domain detection experiment, our method achieves 82.4% F1 value which is about 6% higher than the state-of-the-art method on restaurant domain, and it is also robust on doctor domain.

Improved semi-supervised autoencoder for deception detection.

Existing algorithms of speech-based deception detection are severely restricted by the lack of sufficient number of labelled data. However, a large amount of easily available unlabelled data has not been utilized in reality. To solve this problem, this paper proposes a semi-supervised additive noise autoencoder model for deception detection. This model updates and optimizes the semi-supervised autoencoder and it consists of two layers of encoder and decoder, and a classifier. Firstly, it changes the activation function of the hidden layer in network according to the characteristics of the deception speech. Secondly, in order to prevent over-fitting during training, the specific ratio dropout is added at each layer cautiously. Finally, we directly connected the supervised classification task in the output of encoder to make the network more concise and efficient. Using the feature set specified by the INTERSPEECH 2009 Emotion Challenge, the experimental results on Columbia-SRI-Colorado (CSC) corpus and our own deception corpus show that the proposed model can achieve more advanced performance than other alternative methods with only a small amount of labelled data.

Airport artificial intelligence can detect deception: or am i lying?

Since the 9/11 terrorist attacks, research has enveloped numerous areas within the psychological sciences as a means to increase the ability to spot potential threats. While airports took to heightened security protocols, many academics looked deeper into ways of detecting deception within international airport settings. Various verbal and nonverbal systems were intensely scrutinised under the empirical magnifying glass with the aim of creating security environments that are better able to detect potential threats. However, in 2018, a €4.5 m grant from the European Union’s Horizon 2020 research and innovation programme, number 700,626, was awarded to further in vivo test the use of computational methods to detect deception from facial cues. The system is deemed a noninvasive psychological profiling system and stems from that of a system called ‘Silent Talker’ (Rothwell et al. in Appl Cognit Psychol 20(6):757–777, 2006). The ‘iBorderCtrl’ AI system uses a variety of ‘at home’ pre-registration systems and real time ‘at the airport’ automatic deception detection systems. Some of the critical methods used in automated deception detection are that of micro-expressions. In this opinion article, we argue that considering the state of the psychological sciences current understanding of micro-expressions and their associations with deception, such in vivo testing is naive and misinformed. We consider the lack of empirical research that supports the use of micro-expressions in the detection of deception and question the current understanding of the validity of specific cues to deception. With such unclear definitive and reliable cues to deception, we question the validity of using artificial intelligence that includes cues to deception, which have no current empirical support.

The effect of statement type and repetition on deception detection

BackgroundDeception is a prevalent component of human interaction. However, meta-analyses suggest that discriminating between truthful and deceptive statements is a very arduous task and accuracy on these judgments is at chance levels. To complicate matters further, individuals tell different types of lies. The current studies examined how an individual’s ability to accurately discriminate between truthful and deceptive statements is affected by the way truths and lies are conveyed. Participants judged the veracity of statements given by speakers who told truths or lies about a performed action by describing that action or denying that it had occurred. Additionally, these statements also differed with regard to how often the lie had been repeated (i.e., practiced), either once or thrice.ResultsThe results were largely in line with the prevailing notion that it is difficult to successfully differentiate between truthful and deceptive statements, but also showed that performance was moderated by statement type and repetition. The results revealed that participants were more accurate in discriminating unrepeated descriptions than repeated descriptions, but this difference was not seen for denial statements. Additionally, participants were more likely to believe practiced (repeated) statements, both truthful and deceptive.ConclusionThe results show that repeated statements as well as shorter denials can increase the difficulty of differentiating truthful from deceptive statements. Additionally, these findings suggest that truthful statements also benefit from repetition with regard to enhancing their believability.

A sheep in wolf's clothing: levels of deceit and detection in the evolution of cue-mimicry.

In an evolutionary context, trusted signals or cues provide individuals with the opportunity to manipulate them to their advantage by deceiving others. The deceived can then respond to the deception by either ignoring the signals or cues or evolving means of deception-detection. If the latter happens, it can result in an arms race between deception and detection. Here, we formally analyse these possibilities in the context of cue-mimicry in prey-predator interactions. We demonstrate that two extrinsic parameters control whether and for how long an arms race continues: the benefits of deception, and the cost of ignoring signals and cues and having an indiscriminate response. As long as the cost of new forms of deception is less than its benefits and the cost of new forms of detection is less than the cost of an indiscriminate response, an arms race results in the perpetual evolution of better forms of detection and deception. When novel forms of deception or detection become too costly to evolve, the population settles on a polymorphic equilibrium involving multiple strategies of deception and honesty, and multiple strategies of detection and trust.

Discriminating deception from truth and misinformation: an intent-level approach

ABSTRACTDeception detection has been studied for hundreds of years. A particularly challenging problem is to not only identify truth from deception, but also discriminate misinformation, i.e. errors, from deception. Misinformation has generally been ignored in the study of deception detection, but through analysing the foundations of deception, it may be possible to pinpoint a fundamental difference between deception and all other benign communications – namely, the intent of the speaker. We present a detection model that captures a speaker’s intent by measuring his patterns of reasoning. The reasoning patterns of deceivers may serve as indicators of intentional deception. Through empirical studies, these intent-driven reasoning patterns can identify as well as explain deceptive communications.

Five Reasons Why I Am Skeptical That Indirect or Unconscious Lie Detection Is Superior to Direct Deception Detection.

The relative advantage of indirect and unconscious lie detection compared to direct detection is examined. Empirical evidence for the superiority of indirect and unconscious lie is unconvincing. Three empirical issues include comparisons of incommensurate outcomes, questionable results in control conditions, and evidence for improved performance of direct detection under some conditions. Two theoretical reasons for skepticism include consideration of the casual forces producing poor accuracy and the tendency for people to believe other people absent active cognitive processing. Generally speaking, in human lie detection, effortful and disciplined thought provides more accurate detection of lies than intuition or less than fully conscious cognitive processing.