Deception detection through automatic, unobtrusive analysis of nonverbal behavior

Every day, hundreds of thousands of people pass through airport security checkpoints, border crossing stations, or other security screening measures. Security professionals must sift through countless interactions and ferret out high-risk individuals who represent a danger to other citizens. During each interaction, the security professional must decide whether the individual is being forthright or deceptive. This task is difficult because of the limits of human vigilance and perception and the small percentage of individuals who actually harbor hostile intent. Our research initiative is based on a behavioral approach to deception detection. We attempted to build an automated system that can infer deception or truthfulness from a set of features extracted from head and hands movements in a video. A validated and reliable behaviorally based deception analysis system could potentially have great impacts in augmenting humans' abilities to assess credibility. An automated, unobtrusive system identifies behavioral patterns that indicate deception from nonverbal behavioral cues and classifies deception and truth more accurately than many humans.     

Advancing the assessment of automated deception detection systems: Incorporating base rate and cost into system evaluation

In the last two decades, there has been an increased interest in automated deception detection systems (ADDs) for use in screening, although little attention has been paid to the usefulness of these systems. ADDs use various means, both invasive and non‐invasive, to ascertain individual intent to deceive or engage in malicious behaviour. Many papers introducing ADDs use signal detection theory to compare a technique's ability to detect malicious intent with other techniques, but in doing so, they do not include contextual information such as base rate and cost. In this paper, we aim to improve future research by showing how the inclusion of contextual information provides a more realistic picture of the research. Through both theoretical arguments and a real‐data example, we show that especially for those contexts where malicious intent is infrequent (ie, with low base rates of deception) that not factoring in the base rate overestimates the accuracy and therefore usefulness. We conclude with recommendations for how future research should provide a fuller picture of the accuracy and usefulness of ADDs.     

Cultural determinants of media choice for deception

In today’s business environment, deception is commonplace ( and ). Historically, individuals were limited in their media options, however recent technological advances have given individuals more ways in which to communicate and deceive. The use of these new media change the communication dynamic substantially. Previous studies have found that deception detection differs across media, as media vary in their ability to transmit cues, convey emotion and reversibility, among others. Researchers are beginning to understand the dynamics between media characteristics, individual characteristics and media choice in a deceptive communication context.     

An Introduction to Interactive Discourse Processing from the Perspective of Plan Recognition and Text Planning

People engage in task-oriented dialogues to carry out or plan a task. Each participant in such an interaction must be capable of processing plans in two ways. First, each participant must be capable of understanding the plans that the other participant is using. Researchers have developed theories and models about how computational systems should go about recognizing the plans and goals of another participant, both at the subject-matter level and at the level of the communication. This area of research is called plan recognition. Secondly, each participant must be able to make their owns plans to communicate. This area of natural language research is called text planning.Interactive systems -- systems that understand natural language and that can produce natural language to engage in a task-related interaction -- must address the issue of how understanding plans (the process of plan recognition) relates to making plans for the interaction (the process of text planning). We provide an introduction to these two research areas in natural language processing. Those who need to be familiar with both areas -- to conduct research at their intersection -- will find this introduction useful for building systems that both understand what people are trying to do when they speak and that can actively participate in the interaction.     

Leverage knowledge graph and GCN for fine-grained-level clickbait detection

World Wide Web - Clickbait is the use of an enticing title as bait to deceive users to click. However, the corresponding content is often disappointing, infuriating or even deceitful. This practice...     

Everyday deception or a few prolific liars? The prevalence of lies in text messaging

With the recent and dramatic changes to communication patterns introduced by new information technologies it is increasingly important to understand how deception is produced in new media. In the present study we investigate deception production in text messaging, focusing on how often people lie, about what and to whom. This study uses a novel data collection method that allows for the examination of individuals’ communication records at the message level, which may provide a more accurate account of deception behavior than diary or survey methods. We find that the majority of our participants practiced deception in text messaging. Although lying was a relatively infrequent occurrence for the majority of our participants, there were a small number of prolific liars who told a disproportionately large number of lies using this medium. Additionally, we found some support for the argument that deception occurs less frequently in closer relationships, and we observed how the micro-coordination goals of text messaging change the properties of deceptive text messages relative to face-to-face lies.     

A Statistical Language Modeling Approach to Online Deception Detection

Online deception is disrupting our daily life, organizational process, and even national security. Existing approaches to online deception detection follow a traditional paradigm by using a set of cues as antecedents for deception detection, which may be hindered by ineffective cue identification. Motivated by the strength of statistical language models (SLMs) in capturing the dependency of words in text without explicit feature extraction, we developed SLMs to detect online deception. We also addressed the data sparsity problem in building SLMs in general and in deception detection in specific using smoothing and vocabulary pruning techniques. The developed SLMs were evaluated empirically with diverse datasets. The results showed that the proposed SLM approach to deception detection outperformed a state-of-the-art text categorization method as well as traditional feature-based methods.     

Deception in cyberspace: A comparison of text-only vs. avatar-supported medium

The use of anthropomorphic avatars provides Internet users the opportunity and freedom to manipulate their identity. As cyberspace becomes a haven for deceptive behavior, human–computer interaction research will need to be carried out to study and understand these deceptive behaviors. The objective of this research is to investigate the behavior of deceivers and non-deceivers (or truth-tellers) in the cyberspace environment. We examine if the intention to deceive others influences one's choice of avatars in the online chat environment. We also investigate if communication medium (text-only vs. avatar-supported chat) influences one's perception of trustworthiness of the communication partner. A lab experiment was conducted in an online chat environment with dyads. The results indicate that in the text-only chat environment, subjects who were deceiving their partner experienced higher anxiety levels than those who were truthful to their partner; however, the same phenomenon was not observed in the avatar-supported chat environment. This suggests that “wearing a mask” in cyberspace may reduce anxiety in deceiving others. Additionally, deceivers are more likely to choose avatars that are different from their real selves. The results also show that the use of avatars in a computer-mediated chat environment does not have an impact on one's perceived trustworthiness.     

DOBNet: exploiting the discourse of deception behaviour to uncover online deception strategies

Online deception is fuelled by the escalated penetration of the Internet and social media. As the threat of online deception increases, understanding deception behaviour and underlying strategies is having a greater social impact. The verbal behaviour of online deception has recently been extended to the discourse level; nevertheless, discourse behaviours have been examined in isolation without referring to other behaviours in the discourse. By conceptualising the discourse of online behaviour as a social network (DOBNet), this research investigates possible impacts of deception intent on the central structures of DOBNet at three different levels: the discourse behaviour, subnetwork, and whole network. The empirical results of discourse network analysis and statistical tests provide partial support for each of the hypothesised effects. The findings not only demonstrate the efficacy of discourse in distinguishing deceivers from truth-tellers but also extend deception theories by confirming deception strategies from the perspective of discourse network and by uncovering unique characteristics of online deception strategies.     

A real-time interactive nonverbal communication system through semantic feature extraction as an interlingua

There has been a growing interest in the use of networked virtual environment (NVE) technology to implement telepresence that allows participants to interact with each other in shared cyberspace. In addition, nonverbal language has attracted increased attention because of its association with more natural human communication, and especially sign languages play an important role for the hearing impaired. This paper proposes a novel real-time nonverbal communication system by introducing an artificial intelligence method into the NVE. We extract semantic information as an interlingua from the input text through natural language processing, and then transmit this semantic feature extraction (SFE) to the three-dimensional (3-D) articulated humanoid models prepared for each client in remote locations. Once the SFE is received, the virtual human is animated by the synthesized SFE. Experiments with Japanese and Chinese sign languages show this system makes the real-time animation of avatars available for the participants when chatting with each other. The communication is more natural since it is not just based on text or predefined gesture icons. This proposed system is suitable for sign language distance training as well.     

Text-aware balloon extraction from manga

Manga, a Japanese word for comics, is a worldwide popular visual entertainment. Nowadays, electronic devices boost the fast development of motion manga for the purpose of visual richness and manga promotion. To convert static manga images into motion mangas, text balloons are usually animated individually for better story telling. This needs the artists to cut out each text balloon meticulously, and therefore it is quite labor-intensive and time-consuming. In this paper, we propose an automatic approach that can extract text balloons from manga images both accurately and effectively. Our approach starts by extracting white areas that contain texts as text blobs. Different from existing text blob extraction methods that only rely on shape properties, we incorporate text properties in order to differentiate text blobs from texture blobs. Instead of heuristic parameter thresholding, we achieve text blob classification via learning-based classifiers. Along with the novel text blob classification method, we also make the first attempt in trying to tackle the boundary issue in balloon extraction. We apply our method on various styles of mangas and comics with texts in different languages, and convincing results are obtained in all cases.     

Learning new programming languages: an analysis of the process and problems encountered

Experienced programmers transferring to a new language have a far easier time than the novice learning a first language. However, they still experience considerable difficulties. The objective in this exploratory study was to characterize the kinds of learning and transfer that take place in the early stages of using a new programming language and where difficulties develop. 'Think-aloud' protocols were videotaped as subjects went about trying to write a program in a new programming language. Subjects used One of two unfamiliar languages, one similar to their known language (Pascal) and the other dissimilar. Three types of analyses were done on the recorded protocols: a procedural analysis showing the activities the subjects engaged in as they learned the new language, a programming knowledge analysis showing in which areas of program development difficulties were encountered, and a solutions analysis showing how successful the programmers were at using unique features of the new language. We found that the procedures programmers used to learn a new language were independent of the language being learned. The slight differences that existed in procedures were between levels of expertise. Programmers spent the majority of their time reading a language textbook. The programming knowledge analysis showed that programmers main area of concentration was planning how to implement their approach given the constructs available in the language. We observed many iterations of programmers trying to implement plans, failing and having to revise their plans. Examination of the subjects solutions and implementation approaches in Pascal led us to believe that programmers learning a new language are often biased by their implementation of algorithms in previous languages.     

Learning new programming languages: an analysis of the process and problems encountered

Abstract

Experienced programmers transferring to a new language have a far easier time than the novice learning a first language. However, they still experience considerable difficulties. The objective in this exploratory study was to characterize the kinds of learning and transfer that take place in the early stages of using a new programming language and where difficulties develop. ‘Think-aloud’ protocols were videotaped as subjects went about trying to write a program in a new programming language. Subjects used One of two unfamiliar languages, one similar to their known language (Pascal) and the other dissimilar. Three types of analyses were done on the recorded protocols: a procedural analysis showing the activities the subjects engaged in as they learned the new language, a programming knowledge analysis showing in which areas of program development difficulties were encountered, and a solutions analysis showing how successful the programmers were at using unique features of the new language. We found that the procedures programmers used to learn a new language were independent of the language being learned. The slight differences that existed in procedures were between levels of expertise. Programmers spent the majority of their time reading a language textbook. The programming knowledge analysis showed that programmers main area of concentration was planning how to implement their approach given the constructs available in the language. We observed many iterations of programmers trying to implement plans, failing and having to revise their plans. Examination of the subjects solutions and implementation approaches in Pascal led us to believe that programmers learning a new language are often biased by their implementation of algorithms in previous languages.     

基于校园网的用户追踪技术

为了解决校园网用户信息的记录和跟踪,以便对各种网络安全事件进行自动定位和分析,本文分析了IP地址追踪方法,如何建立攻击路径,以识破网络攻击者的IP地址欺骗等解决校园网用户上网自动记录和追踪的关键技术。     

Verbal irony use in personal blogs

Blogs are a widely growing form of computer-mediated communication used to achieve various personal and professional communicative goals. In the present study, we examined previously posted entries from 71 regular bloggers. We examined the blogs for the use of five forms of verbal irony: hyperbole, understatement, rhetorical question, sarcasm and jocularity. In addition, topic and emotional valence of the ironic utterances were examined. Results showed that hyperbole and understatement were more frequently used than the other forms of ironic language. Discussion of hobbies and social outings was the most commonly occurring topic of ironic language, and bloggers used verbal irony to convey both positive and negative intent. The results of this study demonstrated that adult bloggers do use a variety of forms of verbal irony in their personal blogs, despite the potential risk of being misunderstood.     

Lyrical drama and the “turbid mountebanks”: Styles of dialogue in romantic and renaissance tragedy

Critics have condemned English Romantic tragedies as a series of poor imitations of Renaissance tragedy. This paper tests such “literary-critical” questions through statistical comparisons of ten plays from each group. The measures chosen give evidence of a strong and consistent difference between the groups, going beyond historical changes in the language. The Romantic tragedies are more expository; the Renaissance ones include more commonplace interactions between characters. The later plays do not show the marked variations in function-word frequencies of their predecessors. Of the Renaissance plays, Shakespeare's show the closest affinity to the Romantic tragedies, and the most telling contrasts.     

基于Linux共享库注射技术的网络诱骗系统设计

网络诱骗是一种主动的网络安全防御技术,通过跟踪、监视网络入侵者的行为来分析入侵者的攻击意图和行为特征,从而掌握最新的安全技术.其中,如何捕获入侵者的数据是实施网络诱骗的关键问题.分析了现有的网络诱骗系统的数据捕获方法的不足,并在深入研究Linux共享库注射技术的基础上,提出了一种以该技术为核心的网络诱骗数据捕获方法,使得诱骗系统更难于被发现,提高了诱骗的质量.并结合其它技术,给出了一个以该方法为核心的诱骗系统的实现.     

基于Bi-GRU并包含注意力机制的文本数据真值发现

针对传统真值发现算法无法直接应用于文本数据的问题,该文提出基于Bi-GRU并包含注意力机制的文本数据真值发现方法。根据文本答案的多因素性,词语使用的多样性与文本数据的稀疏性等特点,该文对用户答案进行细粒度划分,并利用Bi-GRU表征文本答案的语义信息,利用双层注意力机制分别学习用户答案关键词可靠度及用户答案可靠度。依据真值发现的一般假设,无监督学习上下文向量,并最终获得可靠答案。实验结果表明,该算法适用于文本数据真值发现场景,较基于检索的方法及传统真值发现算法效果更优。     

基于上下文信息的口语意图检测方法

近年来,随着人工智能的发展与智能设备的普及,人机智能对话技术得到了广泛的关注。口语语义理解是口语对话系统中的一项重要任务,而口语意图检测是口语语义理解中的关键环节。由于多轮对话中存在语义缺失、框架表示以及意图转换等复杂的语言现象,因此面向多轮对话的意图检测任务十分具有挑战性。为了解决上述难题,文中提出了基于门控机制的信息共享网络,充分利用了多轮对话中的上下文信息来提升检测性能。具体而言,首先结合字音特征构建当前轮文本和上下文文本的初始表示,以减小语音识别错误对语义表示的影响;其次,使用基于层级化注意力机制的语义编码器得到当前轮和上下文文本的深层语义表示,包含由字到句再到多轮文本的多级语义信息;最后,通过在多任务学习框架中引入门控机制来构建基于门控机制的信息共享网络,使用上下文语义信息辅助当前轮文本的意图检测。实验结果表明,所提方法能够高效地利用上下文信息来提升口语意图检测效果,在全国知识图谱与语义计算大会(CCKS2018)技术评测任务2的数据集上达到了88.1%的准确率(Acc值)和88.0%的综合正确率(F1值),相比于已有的方法显著提升了性能。     

Language dominance in interpersonal deception in computer-mediated communication

Dominance is not only a complicated social phenomenon that involves interpersonal dynamics, but also an effective strategy used in various applications such as deception detection, negotiation, and online community. The extensive literature on dominance has primarily focused on the personality traits and socio-biological influence, as well as various nonverbal and paralinguistic behaviors associated with dominance. Nonetheless, language dominance manifested through dynamically acquired linguistic capability and strategies has not been fully investigated. The exploration of language dominance in the context of deception is even rarer. With the increasing use of computer-mediated communication (CMC) in all aspects of modern life, language dominance in CMC has emerged as an important issue. This study examines language dominance in the context of deception via CMC. The experimental results show that deceivers: (1) demonstrate a different trend of language dominance from truthtellers over time; (2) manipulate the level of language dominance by initiating communication with low dominance and gradually increasing the level over the course of interaction, and (3) display higher levels of dominance in terms of some linguistic behaviors than truthtellers. They suggest that in CMC, deceivers not only adjust the level of language dominance more frequently, but also change it more remarkably than truthtellers.