Detecting Deception in On and Off-line Communications

Size: px
Start display at page:

Download "Detecting Deception in On and Off-line Communications"

Transcription

1 Detecting Deception in On and Off-line Communications Eliana Feasley and Wesley Tansey ABSTRACT As online services become increasingly prevalent, detection of deception in electronic communication becomes vital to maintaining the many communities built around mutual trust between strangers (e.g. OkCupid, Quora, Yelp). Automating the detection process via machine learning has shown promise in recent years. This paper presents the results of using support vector machines (SVMs) and probabilistic context-free grammars (PCFGs) to classify text as truthful or deceptive. Experiments are conducted on two wellknown datasets of truthful and detective texts. A novel corpus is also developed via an online interrogation game, yielding additional insights regarding detecting deception in online conversations. The results of these experiments suggest that SVMs are superior to PCFGs at classifying deceptive text, and that the classification of deceptive text is highly dependant upon domain. 1. INTRODUCTION The ability to automatically discern whether an individual is writing deceptively is of critical importance to numerous online companies. Question-answer sites rely on accurate responses to openended questions from users, rating sites depend on honest reviews from patrons of local businesses, and dating sites require sincerity in dating profiles and follow-up messages. For such companies, if the veracity of their content cannot be trusted, the value to the end user is effectively zero. Given the vast size of such data and the speed with which it is being generated, detecting deceptive writing necessitates an automated approach. Previous approaches have focused on either simplified algorithm [2, 1] or purely offline communication [3]. While such techniques are useful in some scenarios such as detecting fake reviews, they do not lend insight into interactive communication. As most websites include the ability to ask follow-up questions and comments, a system trained on a corpus of interrogation data may have more practical significance than one trained on a passive corpus. Most previous work to date has also employed only n-gram and part-ofspeech approaches to modeling deception, potentially missing crucial information captured in the structure of deceptive sentences. To address these shortcomings, we conducted a set of experi- Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Copyright 20XX ACM X-XXXXX-XX-X/XX/XX...$ ments on three corpora of deceptive text. We first replicate a subset of the experiments performed in [3] by training an SVM n-gram classifier on their opinion spam dataset. We then apply the same techniques on the dataset from [2] to achieve state-of-the-art performance detecting deceptive text on that corpus. We also gather and analyze a novel dataset of 250 samples of an online interrogation game we developed, representing the first known statisticallymeaningful collection of human evaluations in deception detection. Finally, we present a PCFG-based classification approach and test its effectiveness on all three datasets. We find that the PCFG model performs well in some scenarios but is outperformed by the SVM model in all cases. This paper makes the following novel contributions: 1. An analysis of the performance of an n-gram SVM classifier on the corpus found in [2]. 2. The Deception interrogation game and its associated dataset of 250 matches. 3. An analysis of the performance of human detectives in the Deception dataset. 4. A PCFG-based approach to deception detection and a comparison of its performance to SVMs. The remainder of this paper is organized as follows. Section 2 presents our problem domain and the algorithms we employ. Section 3 presents our experimental setup and describes the details of the datasets and the Deception game. Section 4 presents the results of our experiments and a brief discussion of insights gained. Section 5 describes related work in the area. Section 6 presents some interesting future directions for our work. Section 7 concludes with a summary of our findings. 2. ALGORITHMS The automated detection of deception requires a system capable of assessing an arbitrary piece of prose as truthful or deceitful. Machine learning classifiers such as SVMs are well suited for such tasks as each piece of text can be considered as being labeled with one of two classes, presenting a straight-forward classification problem. It may also be possible that a linguistic model such as a PCFG can perform well on this task by leveraging more finegrained knowledge about language. We next describe these two approaches and how they are used in our experiments. 2.1 SVMs To maintain consistency with previous work [3], we used the SVM light package with a linear kernel to train our SVM.

2 Deception is played with three players, each with a different role. Detective Create a game. Your goal is to pick a topic question, interrogate the two suspects, and determine which one is lying. Both suspects will try to convince you that they are telling the truth about the answer. You re allowed to ask them up to three questions in a game. Truth Teller Join a game and simply tell the truth. Answer each question honestly and convince the detective to believe you. Liar Join a game and convince the detective that you are the truth teller. Pretend that you actually said the main topic question response. Table 1: The Rules of Deception. 2.2 PCFGs For each corpus, we create a PCFG to model deceptive text (P CF G D ) and another to model truthful text (P CF G T ). This is based on the approach in [4]. Part-of-speech tagging is done via the Python Natural Language Toolkit 1 and the Stanford Parser 2 is used to train the PCFGs. In each dataset, we simply take the maximum of the two PCFGs as the classification. In the Deception dataset, however, the detective knows that one of the two individuals is lying. Thus, we evaluate several additional approaches that leverage this knowledge. Algorithm 2.2 lists the four comparative strategies we evaluate. 3. EXPERIMENTAL SETUP We next describe the details of our experiments, including the existing datasets we analyzed and the game we used to collect our own novel corpus. 3.1 Existing Corpora Two existing, previously-published datasets were leveraged in our experiments. First, the PsychLing dataset [2] contains a collection of five experiments, totaling x samples, where subjects were asked to perform various deceptive tasks such as writing false opinions and lying about their previous actions. Second, the OpSpam dataset [3] contains 800 samples of online hotel reviews, half of which are labeled deceitful. Both datasets were provided upon request by the authors of their respective papers. 3.2 The Deception Game To gather our third dataset, we created a web-based interrogation game called Deception 3. The rules of the game are presented to players as shown in Table 1. Using workers from Amazon Mechanical Turk 4 and odesk 5, and participants from various viral social media sites, we gathered over 800 games, over 280 of which were completed (every player answered all questions, and every detective asked all questions and Condition Probability P(R1) P(R2) P(R3) P(R2 R1) P(R3 R2) P(R3 R1, R2) P( R2 R1) P( R3 R2) P( R3 R1, R2) P(R3 switched) Table 2: Marginal and conditional probabilities for the detective guesses in the Deception game dataset. R# represents the interrogation round number; switched refers to the case when the detective switched their guess from round two. The results show that detectives get more accurate as the interrogation progresses, unless they start out on the wrong path. made judgements.) Due to the viral spreading of the game, some of these matches were not played in good faith or involved participants going off-track. Others involved participants asking questions with a yes or no answer (for instance, Have you seen The Avengers "), which leads to a high probability of both players having the same answer, and neither lying. To improve the quality of the dataset, all games were filtered out if they involved the players or questioner directing profanity at one another, or began with a yes-or-no question that was not elaborated on in the answer. After filtering, the Deception dataset contains 250 usable examples, including 1000 questions and 750 instances each of deception, truth-telling, and detective judgements Methodology For each dataset, we apply 5-fold cross-validation on all models and average the performance. All classification results are measured for precision, recall, and F-score. 4. EXPERIMENTAL RESULTS This section presents the results of the experiments on the three deception datasets: OpSpam, PsychLing, and Deception. As both OpSpam and PsychLing are previously-published datasets, we begin by presenting summary statistics about the Deception dataset. We then present the results of our SVM and PCFG experiments on each dataset. The section concludes with a brief discussion of the insights gained from the experiments. 4.1 Deception Statistics To establish a baseline for accuracy, we first measure the probabilities that a human detective will correctly guess the liar in a given round, shown in Table 2. After only one question, human judges perform only slightly better than random but are more than 58% accurate by the end of the final round. Intuitively, if a judge has correctly guessed the liar in a previous round, they are substantially more likely to guess correctly the next round. Conversely, guessing incorrectly appears to have a similar effect as guessing wrong previously is a strong indicator of making future incorrect judgements. Interestingly, if a detective changes their opinion before the third round, they are actually less likely to guess correctly than the marginal probability. 6 See playdeception.com/finished/vcvdelir for an example of a completed game.

3 Algorithm 1 The four comparative strategies used to evaluate PCFGs in the Deception dataset. Require: P CF G T, P CF G D function PCFG-SUM(response1, response2) score1 P CF G D (response1) + P CF G T (response2) score2 P CF G D (response2) + P CF G T (response1) function PCFG-DIFF(response1, response2) score1 P CF G D (response1) P CF G D (response2) score2 P CF G T (response1) P CF G T (response2) function PCFG-LIE(response1, response2) score1 P CF G D (response1) score2 P CF G D (response2) function PCFG-LIE(response1, response2) score1 P CF G T (response1) score2 P CF G T (response2) function MAXLENGTH(response1, response2) if Length(response1) Length(response2) then

4 Approach Accuracy PCFG Sum 48.6 PCFG Difference 55.3 PCFG T ruth 50.2 PCFG Lie 61.8 Table 5: Results on datasets using PCFGS. In addition to analyzing the content of the suspect responses, three descriptive features were also observed: deletions, response length, and time to completion. Table 3 shows the three descriptive features measured. Deletions were approximated by analyzing snapshots of the current suspect response at 1-second intervals and detecting when the text was shorter than it previously had been. The results suggest that deletions and time were not significantly different for liars and truth tellers, however response length appears to be a potential indicator. These results support previous findings [] that show liars create shorter responses due to higher cognitive load. 4.2 Classifier Comparison Table 4 presents the results of our analysis comparing SVMs and PCFGs on our three datasets. While the PCFG model outperforms the part-of-speech SVM, it slightly underperforms the n- gram SVMs in all three experiments. In the case of the Psych- Ling dataset, the unigram SVM attains the highest known accuracy to-date. Interestingly, the unigram SVM also had the highest accuracy on the Deception dataset, though in both cases the F-score was lower than the bigram and trigram models, respectively. Table 4.2 presents the results of the PCFG variant models on the Deception data. These models have higher accuracy as they are leveraging the knowledge that one of the suspects is lying and the other is telling the truth. It appears that the best performing model is thus the PCFG Lie algorithm. However, it is worth noting that the majority of this prediction accuracy comes from using the length function as a tie-breaker. Using purely length is sufficient to achieve 61.6% accuracy on the final guess. 4.3 Discussion Deception detection has a long way to go. People are good at lying and are well practiced in conveying the impression of telling the truth. Our methods are comparable to those of a human judge, but building a discriminator based on number of words in the total communication was more effective than either natural language processing or human judges. In certian domains, deceptian detection may be easier or harder - for instance, in the OpSpam domain, the authors of one set of documents are Mechanical Turkers while the author of the other set are people who can afford to stay in fivestar hotels in Chicago. The strong influence of domain is evident when comparing the most important features in the SVM model generated for each corpus (See Figure 6). In the bigram model for the PsychLing corpus, the SVM seems to be discriminating morebet In the Psych- Ling dataset, the SVM is discriminating more based on the prolife/pro-choice stance of the student than on any high-level linguistic traits, with phrases classically associated with being pro-life such as kill, murder, and life appearing as features for deception, while the indicators for truthfulness contain more personal pronouns and counterfactuals (the phrases that if and it would ). In contrast, the OpSpam dataset contains many occurances of personal words like i and my as deceptive features, while the truthful accounts contain references to concrete nouns - the hotel, the bathroom, the concierge. These indicators are nearly opposite. In the Deception dataset, there is no intuitive explanation for the differences between the datasets - and indeed, the model performed no better than chance. The PCFG models were not very effective for the Deception dataset. This is likely because the highly informal chat-based format, involving many fragments and one-word responses, is not very amenable to a grammar being generated. The SVMs may have been more effective because they were able to take the entire document into account, rather than one sentence - or fragment - at a time. 5. RELATED WORK Analyzing text to detect deception has received considerable attention in psycholinguistics, but has only recently become the subject of study in the machine learning field. As a result, a majority of previous works rely on bag-of-words analyses and attempt to draw correlations between categories of words and deceptive text. For instance, Newman et al. [2] conducted a set of experiments in which subjects were asked to either lie or tell the truth about a topic. The tasks include presenting an opinion on abortion, with participants either writing, speaking, or typing both their opinion and the reverse of their opinion (pro-life vs. pro-choice); participants describing friends they like and dislike either positively or negatively; and participants either lying or telling the truth under interrogation about whether or not they had stolen a one dollar bill from a book in the middle of a room. The resulting dataset was analyzed via the Linguistic Inquiry and Word Count (LIWC) tool, looking only at individual words and performing logistic regression. While the system slightly outperformed human judges, we have shown that n-gram SVMs are a superior model to both approaches on the same dataset. More directly related, Hancock et al. [1] investigate distinctions between attributes of deceptive communication in online, verbal, and written communication. Some participants were motivated to lie successfully via a lying pep talk detailing the ways in which lying is essential for a successful life, while others were merely told to lie. Transcripts of 242 text-based chats in which participants either lied or told the truth were analyzed. The results support previous work showing that liars are less likely to use first person singular pronouns and negatives, while being more likely to use third person singular. Similarly to Newman et al., no effort is made to go beyond surface-level statistical analysis via a bag-of-words model. Ott et al. [3] present an approach to detecting opinion spam in online hotel reviews. The authors gather the first large-scale (800 total), gold-standard dataset of truthful and deceptive positive reviews, with half being deceptive and half truthful. Four main experiments are conducted to determine the feasibility of detecting deception. First, three human judges attempt to classify each review; results indicate that humans are poor judges and only achieve 60% accuracy. Three automated approaches are then tried: genre identification, psycholinguistic features, and text categorization. Genre identification relies purely on part-of-speech tags to attempt to detect whether the two classes are stylistically different. Psycholinguistic features derived from LIWC are combined with an SVM model to generate a classifier. Finally, n-gram models are combined with SVM and Naive Bayes (NB) to generate another collection of text classifiers. Permutations of LIWC with n-grams are also explored. The results show that n-gram + LIWC categorization methods are the most effective at the task (nearly 90% accuracy), though only slightly (0.02%) more effective than a purely n-gram SVM model, while even the simple genre identification model statistically significantly outperforms humans. The authors show confirmation of previous studies that deceivers have difficulty with spatial

5 Deletions Words Seconds Suspect Mean Median Stdev p-value Truth Teller Liar Truth Teller Liar Truth Teller Liar Table 3: Summary statistics for the suspects in the Deception game dataset. The results support previous findings that liars write less due to a higher cognitive load. Approach Corpus Accuracy P R F TruthAcc LieAcc Unigram SV M OpSpam Bigram SV M OpSpam Trigram SV M OpSpam POS SV M OpSpam PCFG OpSpam Unigram SV M PsychLing Bigram SV M PsychLing Trigram SV M PsychLing POS SV M PsychLing PCFG PsychLing Unigram SV M Deception Bigram SV M Deception Trigram SV M Deception POS SV M Deception PCFG Deception Table 4: Results using n-gram and part-of-speech features with SVMs on the OpSpam and PsychLing datasets. PsychLing PsychLing OpSpam OpSpam Deception Deception that if on the i dont do with we had i think i believe the hotel i guess that she the bathroom it is their own the concierge yes i but i michigan ave of a i can is a dont really it would room was that it have to distance to to get i think michigan avenue i wasnt 0.82 and then 1.03 the area 0.72 i live 0.89 have the 1.04 east hotel 0.76 a great 0.93 to take 1.07 to do 0.76 to explain 1.04 killing a 1.09 the food 0.82 i like 1.08 should not 1.12 next time 0.82 to me 1.13 shouldnt be 1.16 when i 0.84 a lot 1.15 because it 1.23 chicago hotel 0.90 i would 1.27 is murder 1.26 and i 0.93 i get 1.33 a life 1.38 to anyone 0.94 think it 1.47 a human 1.67 my stay 0.97 the way 1.52 to kill 1.75 my husband 1.63 i can Table 6: Weights on features in the linear-kernel SVM. Low weights predict truth; high weights, deception.

6 and temporal reasoning by analyzing the highest weighted words in their final models and noting that truthful reviews contain words like location, on, and bathroom. This same analysis also seems to contradict previous findings by showing that deceptive reviewers use pronouns like I and my frequently. The authors conclude this shows a need to analyze deception within the context of the given task and that a global model is insufficient. We borrow heavily from Ott et al. in this paper, first replicating their results 7 and then extending the technique to the other two datasets. It is worth noting that while the OpSpam dataset is large, it is also an offline dataset that does not include the interrogative components of the Deception dataset. 6. LIMITATIONS AND FUTURE WORK The introduction of the Deception game yields several promising directions for future research. First, the data is not particularly high-quality as the popularity of the site attracted a wide range of players, resulting in a correspondingly wide range of games. To overcome this, we plan to organize a large-scale controlled experiment with undergraduate students. Second, we have not yet done an in-depth analysis of the questions asked by the detective. Such an analysis may reveal insights into which questions yield the most informative answers from the suspects. Similarly, we have captured snapshots of all the text entered by the suspects at every second, but only performed a very surface-level analysis of the typing patterns. It would also be interesting to determine the success of detectives when they cannot see the suspects typing in real-time as opposed to our current data in which they can. 7. CONCLUSION Much work has been done in both psycholinguistics and computational linguistics in identifying deception, but this work has not been fully synthesized in the past. This paper has presented the results of applying a PCFG-based approach to two previously published corpora as well as a novel third corpus. While PCFGs were able to perform well on some of the datasets, they were inferior to SVMs in all three experiments. 8. REFERENCES [1] J.T. Hancock, L.E. Curry, S. Goorha, and M. Woodworth. On lying and being lied to: A linguistic analysis of deception in computer-mediated communication. Discourse Processes, 45(1):1 23, [2] M.L. Newman, J.W. Pennebaker, D.S. Berry, and J.M. Richards. Lying words: Predicting deception from linguistic styles. Personality and Social Psychology Bulletin, 29(5):665, [3] M. Ott, Y. Choi, C. Cardie, and J.T. Hancock. Finding deceptive opinion spam by any stretch of the imagination. Arxiv preprint arxiv: , [4] Sindhu Raghavan, Adriana Kovashka, and Raymond Mooney. Authorship attribution using probabilistic context-free grammars. In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics (ACL-2010), pages 38 42, We were unable to achieve the same level of performance using SVM light and a linear kernel. Nevertheless, we achieved sufficiently high results to allow for adequate comparison.

Motivation. Motivation. Motivation. Finding Deceptive Opinion Spam by Any Stretch of the Imagination

Motivation. Motivation. Motivation. Finding Deceptive Opinion Spam by Any Stretch of the Imagination Finding Deceptive Opinion Spam by Any Stretch of the Imagination Myle Ott, 1 Yejin Choi, 1 Claire Cardie, 1 and Jeff Hancock 2! Dept. of Computer Science, 1 Communication 2! Cornell University, Ithaca,

More information

Negative Deceptive Opinion Spam

Negative Deceptive Opinion Spam Negative Deceptive Opinion Spam Myle Ott Claire Cardie Department of Computer Science Cornell University Ithaca, NY 14853 {myleott,cardie}@cs.cornell.edu Jeffrey T. Hancock Department of Communication

More information

Cross-cultural Deception Detection

Cross-cultural Deception Detection Cross-cultural Deception Detection Verónica Pérez-Rosas Computer Science and Engineering University of North Texas veronicaperezrosas@my.unt.edu Rada Mihalcea Computer Science and Engineering University

More information

DECEPTION DETECTION COMPUTATIONAL APPROACHES

DECEPTION DETECTION COMPUTATIONAL APPROACHES DECEPTION DETECTION COMPUTATIONAL APPROACHES OVERVIEW DEFINING DECEPTION To intentionally cause another person to have or continue to have a false belief that is truly believed to be false by the person

More information

The truth about lying

The truth about lying Reading Practice The truth about lying Over the years Richard Wiseman has tried to unravel the truth about deception - investigating the signs that give away a liar. A In the 1970s, as part of a large-scale

More information

Deception and its detection - A brief overview

Deception and its detection - A brief overview Deception and its detection - A brief overview Franziska Clemens, PhD candidate Department of Psychology University of Gothenburg Overview Deception Deception Detection How good are people at detecting

More information

Psychological profiling through textual analysis

Psychological profiling through textual analysis Psychological profiling through textual analysis... John Noecker Jr, Michael Ryan and Patrick Juola Duquesne University... Abstract In this article, we examine the application of computational stylometry

More information

Case Studies of Signed Networks

Case Studies of Signed Networks Case Studies of Signed Networks Christopher Wang December 10, 2014 Abstract Many studies on signed social networks focus on predicting the different relationships between users. However this prediction

More information

Generalizing Dependency Features for Opinion Mining

Generalizing Dependency Features for Opinion Mining Generalizing Dependency Features for Mahesh Joshi 1 and Carolyn Rosé 1,2 1 Language Technologies Institute 2 Human-Computer Interaction Institute Carnegie Mellon University ACL-IJCNLP 2009 Short Papers

More information

H O W T O T E L L I F S O M E O N E I S LY I N G J O N S T E T S O N

H O W T O T E L L I F S O M E O N E I S LY I N G J O N S T E T S O N H O W T O T E L L I F S O M E O N E I S LY I N G B Y J O N S T E T S O N THE KNOWLEDGE DISCUSSED in this book can be useful for managers, employers, and for anyone to use in everyday situations where telling

More information

Building Evaluation Scales for NLP using Item Response Theory

Building Evaluation Scales for NLP using Item Response Theory Building Evaluation Scales for NLP using Item Response Theory John Lalor CICS, UMass Amherst Joint work with Hao Wu (BC) and Hong Yu (UMMS) Motivation Evaluation metrics for NLP have been mostly unchanged

More information

Lecture 10: POS Tagging Review. LING 1330/2330: Introduction to Computational Linguistics Na-Rae Han

Lecture 10: POS Tagging Review. LING 1330/2330: Introduction to Computational Linguistics Na-Rae Han Lecture 10: POS Tagging Review LING 1330/2330: Introduction to Computational Linguistics Na-Rae Han Overview Part-of-speech tagging Language and Computers, Ch. 3.4 Tokenization, POS tagging NLTK Book Ch.5

More information

FSA Training Papers Grade 7 Exemplars. Rationales

FSA Training Papers Grade 7 Exemplars. Rationales FSA Training Papers Grade 7 Exemplars Rationales Rationales for Grade 7 Exemplars Reading Grade 7 Reading Exemplar #1: Score 3 Comprehension of the passages and task clearly evident Generally purposeful

More information

This week s issue: UNIT Word Generation. conceive unethical benefit detect rationalize

This week s issue: UNIT Word Generation. conceive unethical benefit detect rationalize Word Generation This week s issue: We all know the story about George Washington s honesty. As a little boy George chopped down a cherry tree with his shiny new axe. When confronted by his father he immediately

More information

Sentiment Analysis of Reviews: Should we analyze writer intentions or reader perceptions?

Sentiment Analysis of Reviews: Should we analyze writer intentions or reader perceptions? Sentiment Analysis of Reviews: Should we analyze writer intentions or reader perceptions? Isa Maks and Piek Vossen Vu University, Faculty of Arts De Boelelaan 1105, 1081 HV Amsterdam e.maks@vu.nl, p.vossen@vu.nl

More information

Identifying Signs of Depression on Twitter Eugene Tang, Class of 2016 Dobin Prize Submission

Identifying Signs of Depression on Twitter Eugene Tang, Class of 2016 Dobin Prize Submission Identifying Signs of Depression on Twitter Eugene Tang, Class of 2016 Dobin Prize Submission INTRODUCTION Depression is estimated to affect 350 million people worldwide (WHO, 2015). Characterized by feelings

More information

ALABAMA SELF-ASSESSMENT INDEX PILOT PROGRAM SUMMARY REPORT

ALABAMA SELF-ASSESSMENT INDEX PILOT PROGRAM SUMMARY REPORT ALABAMA SELF-ASSESSMENT INDEX PILOT PROGRAM SUMMARY REPORT July 7, 2000 A large sample (N = 1,127) of welfare recipient clients were administered the Self-Assessment Index Behavior Data Systems, Ltd. P.O.

More information

Running Head: AUTOMATED SCORING OF CONSTRUCTED RESPONSE ITEMS. Contract grant sponsor: National Science Foundation; Contract grant number:

Running Head: AUTOMATED SCORING OF CONSTRUCTED RESPONSE ITEMS. Contract grant sponsor: National Science Foundation; Contract grant number: Running Head: AUTOMATED SCORING OF CONSTRUCTED RESPONSE ITEMS Rutstein, D. W., Niekrasz, J., & Snow, E. (2016, April). Automated scoring of constructed response items measuring computational thinking.

More information

EMOTION CLASSIFICATION: HOW DOES AN AUTOMATED SYSTEM COMPARE TO NAÏVE HUMAN CODERS?

EMOTION CLASSIFICATION: HOW DOES AN AUTOMATED SYSTEM COMPARE TO NAÏVE HUMAN CODERS? EMOTION CLASSIFICATION: HOW DOES AN AUTOMATED SYSTEM COMPARE TO NAÏVE HUMAN CODERS? Sefik Emre Eskimez, Kenneth Imade, Na Yang, Melissa Sturge- Apple, Zhiyao Duan, Wendi Heinzelman University of Rochester,

More information

Classification of deceptive opinions using a low dimensionality representation

Classification of deceptive opinions using a low dimensionality representation Classification of deceptive opinions using a low dimensionality representation Leticia C. Cagnina LIDIC Universidad Nacional de San Luis San Luis, Argentina lcagnina@unsl.edu.ar Paolo Rosso NLE Lab, PRHLT

More information

Gender Differences in Deceivers Writing Style

Gender Differences in Deceivers Writing Style Gender Differences in Deceivers Writing Style Verónica Pérez-Rosas and Rada Mihalcea University of North Texas, University of Michigan vrncapr@umich.edu, mihalcea@umich.edu Abstract. The widespread use

More information

Identifying Personality Types Using Document Classification Methods

Identifying Personality Types Using Document Classification Methods Identifying Personality Types Using Document Classification Methods Mike Komisin and Curry Guinn Department of Computer Science, University of North Carolina Wilmington mkomisin@gmail.com, guinnc@uncw.edu

More information

Assessing credibility

Assessing credibility Assessing credibility Judgecraft across Europe Omsenie, Slovakia 12 October 2015 John Phillips Outline What is credibility? What research is there and what are its limitations? What are the non-verbal

More information

Detecting Hoaxes, Frauds and Deception in Writing Style Online

Detecting Hoaxes, Frauds and Deception in Writing Style Online Detecting Hoaxes, Frauds and Deception in Writing Style Online Sadia Afroz, Michael Brennan and Rachel Greenstadt Privacy, Security and Automation Lab Drexel University What do we mean by deception? Let

More information

Detection of Opinion Spam with Character n-grams

Detection of Opinion Spam with Character n-grams Detection of Opinion Spam with Character n-grams Donato Hernández Fusilier,2, Manuel Montes-y-Gómez 3, Paolo Rosso, and Rafael Guzmán Cabrera 2 Natural Language Engineering Lab., Universitat Politècnica

More information

Good Communication Starts at Home

Good Communication Starts at Home Good Communication Starts at Home It is important to remember the primary and most valuable thing you can do for your deaf or hard of hearing baby at home is to communicate at every available opportunity,

More information

Quoting extensively from another source, even if you do it properly, is not appropriate

Quoting extensively from another source, even if you do it properly, is not appropriate This guide will show you how to use information correctly without plagiarizing. You will learn how to recognize plagiarism, and how to tell the difference between plagiarism and appropriate use of information

More information

Asthma Surveillance Using Social Media Data

Asthma Surveillance Using Social Media Data Asthma Surveillance Using Social Media Data Wenli Zhang 1, Sudha Ram 1, Mark Burkart 2, Max Williams 2, and Yolande Pengetnze 2 University of Arizona 1, PCCI-Parkland Center for Clinical Innovation 2 {wenlizhang,

More information

Modeling Sentiment with Ridge Regression

Modeling Sentiment with Ridge Regression Modeling Sentiment with Ridge Regression Luke Segars 2/20/2012 The goal of this project was to generate a linear sentiment model for classifying Amazon book reviews according to their star rank. More generally,

More information

Political Science 15, Winter 2014 Final Review

Political Science 15, Winter 2014 Final Review Political Science 15, Winter 2014 Final Review The major topics covered in class are listed below. You should also take a look at the readings listed on the class website. Studying Politics Scientifically

More information

Get a Clue! Some Truths about Online Deception

Get a Clue! Some Truths about Online Deception Get a Clue! Some Truths about Online Deception Cheryl Lynn Booth 1, Shuyuan Mary Ho 1 1 Florida State University, School of Information Abstract As text-based computer-mediated technologies have become

More information

THE INTEGRITY PROFILING SYSTEM

THE INTEGRITY PROFILING SYSTEM THE INTEGRITY PROFILING SYSTEM The Integrity Profiling System which Soft Skills has produced is the result of a world first research into leadership and integrity in Australia. Where we established what

More information

AP Psychology -- Chapter 02 Review Research Methods in Psychology

AP Psychology -- Chapter 02 Review Research Methods in Psychology AP Psychology -- Chapter 02 Review Research Methods in Psychology 1. In the opening vignette, to what was Alicia's condition linked? The death of her parents and only brother 2. What did Pennebaker s study

More information

5. is the process of moving from the specific to the general. a. Deduction

5. is the process of moving from the specific to the general. a. Deduction Applied Social Psychology Understanding and Addressing Social and Practical Problems 3rd Edition Gruman Test Bank Full Download: https://testbanklive.com/download/applied-social-psychology-understanding-and-addressing-social-and-practical-p

More information

arxiv: v2 [cs.ai] 26 Sep 2018

arxiv: v2 [cs.ai] 26 Sep 2018 Manipulating and Measuring Model Interpretability arxiv:1802.07810v2 [cs.ai] 26 Sep 2018 Forough Poursabzi-Sangdeh forough.poursabzi@microsoft.com Microsoft Research Jennifer Wortman Vaughan jenn@microsoft.com

More information

Classification of Honest and Deceitful Memory in an fmri Paradigm CS 229 Final Project Tyler Boyd Meredith

Classification of Honest and Deceitful Memory in an fmri Paradigm CS 229 Final Project Tyler Boyd Meredith 12/14/12 Classification of Honest and Deceitful Memory in an fmri Paradigm CS 229 Final Project Tyler Boyd Meredith Introduction Background and Motivation In the past decade, it has become popular to use

More information

Modeling the Use of Space for Pointing in American Sign Language Animation

Modeling the Use of Space for Pointing in American Sign Language Animation Modeling the Use of Space for Pointing in American Sign Language Animation Jigar Gohel, Sedeeq Al-khazraji, Matt Huenerfauth Rochester Institute of Technology, Golisano College of Computing and Information

More information

Author s Traits Prediction on Twitter Data using Content Based Approach

Author s Traits Prediction on Twitter Data using Content Based Approach Author s Traits Prediction on Twitter Data using Content Based Approach Notebook for PAN at CLEF 2015 Fahad Najib, Waqas Arshad Cheema, Rao Muhammad Adeel Nawab Department of Computer Science, COMSATS

More information

Rumor Detection on Twitter with Tree-structured Recursive Neural Networks

Rumor Detection on Twitter with Tree-structured Recursive Neural Networks 1 Rumor Detection on Twitter with Tree-structured Recursive Neural Networks Jing Ma 1, Wei Gao 2, Kam-Fai Wong 1,3 1 The Chinese University of Hong Kong 2 Victoria University of Wellington, New Zealand

More information

Studying the Dark Triad of Personality through Twitter Behavior

Studying the Dark Triad of Personality through Twitter Behavior Studying the Dark Triad of Personality through Twitter Behavior Daniel Preoţiuc-Pietro Jordan Carpenter, Salvatore Giorgi, Lyle Ungar Positive Psychology Center Computer and Information Science University

More information

Cultural Competence: An Ethical Model for Big Data Research

Cultural Competence: An Ethical Model for Big Data Research Cultural Competence: An Ethical Model for Big Data Research Amirah Majid The Information School University of Washington Seattle, WA 98105 USA amirah@uw.edu Abstract Data science research is human subjects

More information

Evaluating Classifiers for Disease Gene Discovery

Evaluating Classifiers for Disease Gene Discovery Evaluating Classifiers for Disease Gene Discovery Kino Coursey Lon Turnbull khc0021@unt.edu lt0013@unt.edu Abstract Identification of genes involved in human hereditary disease is an important bioinfomatics

More information

Finding Deceptive Opinion Spam by Any Stretch of the Imagination

Finding Deceptive Opinion Spam by Any Stretch of the Imagination Finding Deceptive Opinion Spam by Any Stretch of the Imagination Myle Ott Yejin Choi Claire Cardie Department of Computer Science Cornell University Ithaca, NY 14853 {myleott,ychoi,cardie}@cs.cornell.edu

More information

the ebook on To learn more about human dynamics, grab your copy of: Sales is a People Game. PEOPLE RICH PUBLISHING Australian Business Number: 40 137 075 435 PO Box 39 NORTHLANDS WA 6905 AUSTRALIA All

More information

Situation Reaction Detection Using Eye Gaze And Pulse Analysis

Situation Reaction Detection Using Eye Gaze And Pulse Analysis Situation Reaction Detection Using Eye Gaze And Pulse Analysis 1 M. Indumathy, 2 Dipankar Dey, 2 S Sambath Kumar, 2 A P Pranav 1 Assistant Professor, 2 UG Scholars Dept. Of Computer science and Engineering

More information

Looking for Subjectivity in Medical Discharge Summaries The Obesity NLP i2b2 Challenge (2008)

Looking for Subjectivity in Medical Discharge Summaries The Obesity NLP i2b2 Challenge (2008) Looking for Subjectivity in Medical Discharge Summaries The Obesity NLP i2b2 Challenge (2008) Michael Roylance and Nicholas Waltner Tuesday 3 rd June, 2014 Michael Roylance and Nicholas Waltner Looking

More information

CREDIBILITY ASSESSMENT: PITFALLS AND OPPORTUNITIES. Dr Lucy Akehurst University of Portsmouth Psychology Department

CREDIBILITY ASSESSMENT: PITFALLS AND OPPORTUNITIES. Dr Lucy Akehurst University of Portsmouth Psychology Department CREDIBILITY ASSESSMENT: PITFALLS AND OPPORTUNITIES Dr Lucy Akehurst University of Portsmouth Psychology Department Email: lucy.akehurst@port.ac.uk CUES TO DECEIT What are nonverbal cues? Based on the research

More information

When Twitter meets Foursquare: Tweet Location Prediction using Foursquare

When Twitter meets Foursquare: Tweet Location Prediction using Foursquare When Twitter meets Foursquare: Tweet Location Prediction using Foursquare Kisung Lee, Raghu K. Ganti, Mudhakar Srivatsa, Ling Liu College of Computing, Georgia Institute of Technology, Atlanta, GA USA

More information

Use of Twitter to Assess Sentiment toward Waterpipe Tobacco Smoking

Use of Twitter to Assess Sentiment toward Waterpipe Tobacco Smoking @ColditzJB #SBM2016 Use of Twitter to Assess Sentiment toward Waterpipe Tobacco Smoking Jason B. Colditz, MEd Maharsi Naidu, Class of 2018 Noah A. Smith, PhD Joel Welling, PhD Brian A. Primack, MD, PhD

More information

Critical Conversations

Critical Conversations Critical Conversations TIPS FOR TALKING WHEN STAKES ARE HIGH Agenda 1. Basics of Communication 2. Crucial Conversations defined 3. Before the conversation: setting the stage 4. During the conversation:

More information

Social and Pragmatic Language in Autistic Children

Social and Pragmatic Language in Autistic Children Parkland College A with Honors Projects Honors Program 2015 Social and Pragmatic Language in Autistic Children Hannah Li Parkland College Recommended Citation Li, Hannah, "Social and Pragmatic Language

More information

Third Meditation: Imperfect Advice

Third Meditation: Imperfect Advice 10 Third Meditation: Imperfect Advice In my early professional years I was asking the question: How can I treat, or cure, or change this person? Now I would phrase the question in this way: How can I provide

More information

How to Create Better Performing Bayesian Networks: A Heuristic Approach for Variable Selection

How to Create Better Performing Bayesian Networks: A Heuristic Approach for Variable Selection How to Create Better Performing Bayesian Networks: A Heuristic Approach for Variable Selection Esma Nur Cinicioglu * and Gülseren Büyükuğur Istanbul University, School of Business, Quantitative Methods

More information

Identify and leverage your most powerful influencing skills. Date. Name. Organization Name

Identify and leverage your most powerful influencing skills. Date. Name. Organization Name Identify and leverage your most powerful influencing skills Name Date Organization Name Introduction ability to influence and persuade others is key to success. People who lack influencing skills tend

More information

How Self-Efficacy and Gender Issues Affect Software Adoption and Use

How Self-Efficacy and Gender Issues Affect Software Adoption and Use How Self-Efficacy and Gender Issues Affect Software Adoption and Use Kathleen Hartzel Today s computer software packages have potential to change how business is conducted, but only if organizations recognize

More information

ACCESS TO DENTAL CARE FOR BRISTOL STUDENTS

ACCESS TO DENTAL CARE FOR BRISTOL STUDENTS ACCESS TO DENTAL CARE FOR BRISTOL STUDENTS Healthwatch engaged with students studying at Bristol University to find out students experiences of using local dental services. The questionnaire aimed to identify

More information

Text mining for lung cancer cases over large patient admission data. David Martinez, Lawrence Cavedon, Zaf Alam, Christopher Bain, Karin Verspoor

Text mining for lung cancer cases over large patient admission data. David Martinez, Lawrence Cavedon, Zaf Alam, Christopher Bain, Karin Verspoor Text mining for lung cancer cases over large patient admission data David Martinez, Lawrence Cavedon, Zaf Alam, Christopher Bain, Karin Verspoor Opportunities for Biomedical Informatics Increasing roll-out

More information

Exploring Normalization Techniques for Human Judgments of Machine Translation Adequacy Collected Using Amazon Mechanical Turk

Exploring Normalization Techniques for Human Judgments of Machine Translation Adequacy Collected Using Amazon Mechanical Turk Exploring Normalization Techniques for Human Judgments of Machine Translation Adequacy Collected Using Amazon Mechanical Turk Michael Denkowski and Alon Lavie Language Technologies Institute School of

More information

Deception Detection Accuracy Using Verbal or Nonverbal Cues

Deception Detection Accuracy Using Verbal or Nonverbal Cues The Journal of Undergraduate Research Volume 9 Journal of Undergraduate Research, Volume 9: 2011 Article 9 2011 Deception Detection Accuracy Using Verbal or Nonverbal Cues Caroline Hicks South Dakota State

More information

Introduction to Sentiment Analysis

Introduction to Sentiment Analysis Introduction to Sentiment Analysis Machine Learning and Modelling for Social Networks Lloyd Sanders, Olivia Woolley, Iza Moize, Nino Antulov-Fantulin D-GESS: Computational Social Science Overview What

More information

Gender-based Multimodal Deception Detection

Gender-based Multimodal Deception Detection Gender-based Multimodal Deception Detection Mohamed Abouelenien Computer Science and Engineering University of Michigan Ann Arbor, MI 48109, USA zmohamed@umich.edu Rada Mihalcea Computer Science and Engineering

More information

Cambridge International AS & A Level Global Perspectives and Research. Component 4

Cambridge International AS & A Level Global Perspectives and Research. Component 4 Cambridge International AS & A Level Global Perspectives and Research 9239 Component 4 In order to help us develop the highest quality Curriculum Support resources, we are undertaking a continuous programme

More information

The! Lie Detection Cheat Sheet!

The! Lie Detection Cheat Sheet! The Lie Detection Cheat Sheet By Paul Mascetta Welcome to the The Lie Detection Cheat Sheet. I realize this can be a touchy or even controversial topic. Listen, I m a firm believer in the power of positive

More information

More skilled internet users behave (a little) more securely

More skilled internet users behave (a little) more securely More skilled internet users behave (a little) more securely Elissa Redmiles eredmiles@cs.umd.edu Shelby Silverstein shelby93@umd.edu Wei Bai wbai@umd.edu Michelle L. Mazurek mmazurek@umd.edu University

More information

Commentary on The Erotetic Theory of Attention by Philipp Koralus. Sebastian Watzl

Commentary on The Erotetic Theory of Attention by Philipp Koralus. Sebastian Watzl Commentary on The Erotetic Theory of Attention by Philipp Koralus A. Introduction Sebastian Watzl The study of visual search is one of the experimental paradigms for the study of attention. Visual search

More information

Experimental evidence of massive-scale emotional contagion through social networks

Experimental evidence of massive-scale emotional contagion through social networks Experimental evidence of massive-scale emotional contagion through social networks September 26, 2016 Goal of Experiment Problems to be solved in Experiment Details of Experiment Conclusions Experiment

More information

Collaboration in neurodivergent teams

Collaboration in neurodivergent teams Collaboration in neurodivergent teams March 2018 Neurodiversity Life Goal social conformance knowledge action art power 1. Neurotypical Disorder 90% (roughly) 2. Autism Spectrum Disorder(s) 1.5% (1 in

More information

UniNE at CLEF 2015: Author Profiling

UniNE at CLEF 2015: Author Profiling UniNE at CLEF 2015: Author Profiling Notebook for PAN at CLEF 2015 Mirco Kocher University of Neuchâtel rue Emile Argand 11 2000 Neuchâtel, Switzerland Mirco.Kocher@unine.ch Abstract. This paper describes

More information

Exploiting Ordinality in Predicting Star Reviews

Exploiting Ordinality in Predicting Star Reviews Exploiting Ordinality in Predicting Star Reviews Alim Virani UBC - Computer Science alim.virani@gmail.com Chris Cameron UBC - Computer Science cchris13@cs.ubc.ca Abstract Automatically evaluating the sentiment

More information

Module 14: Missing Data Concepts

Module 14: Missing Data Concepts Module 14: Missing Data Concepts Jonathan Bartlett & James Carpenter London School of Hygiene & Tropical Medicine Supported by ESRC grant RES 189-25-0103 and MRC grant G0900724 Pre-requisites Module 3

More information

COGNITIVE DISTORTIONS

COGNITIVE DISTORTIONS COGNITIVE DISTORTIONS Cognitive distortions are simply ways that Impostor Syndrome convinces us to believe things that aren t really true. These are inaccurate thought patterns that reinforce our negative

More information

Introduction to Research Methods

Introduction to Research Methods Introduction to Research Methods Updated August 08, 2016 1 The Three Types of Psychology Research Psychology research can usually be classified as one of three major types: 1. Causal Research When most

More information

Cannabis. Screening and Action Planning Toolkit. A toolkit for those who are concerned about their cannabis use and those who support them.

Cannabis. Screening and Action Planning Toolkit. A toolkit for those who are concerned about their cannabis use and those who support them. Cannabis Screening and Action Planning Toolkit A toolkit for those who are concerned about their cannabis use and those who support them. V1.: 015 About this tool: Cannabis dependency hasn t always been

More information

Consulting Skills. Part 1: Critical assessment of Peter Block and Edgar Schein s frameworks

Consulting Skills. Part 1: Critical assessment of Peter Block and Edgar Schein s frameworks Consulting Skills Part 1: Critical assessment of Peter Block and Edgar Schein s frameworks Anyone with their sights set on becoming a consultant or simply looking to improve their existing consulting skills

More information

The Long Tail of Recommender Systems and How to Leverage It

The Long Tail of Recommender Systems and How to Leverage It The Long Tail of Recommender Systems and How to Leverage It Yoon-Joo Park Stern School of Business, New York University ypark@stern.nyu.edu Alexander Tuzhilin Stern School of Business, New York University

More information

Prediction of Average and Perceived Polarity in Online Journalism

Prediction of Average and Perceived Polarity in Online Journalism Prediction of Average and Perceived Polarity in Online Journalism Albert Chu, Kensen Shi, Catherine Wong Abstract We predicted the average and perceived journalistic objectivity in online news articles

More information

The Lens Model and Linear Models of Judgment

The Lens Model and Linear Models of Judgment John Miyamoto Email: jmiyamot@uw.edu October 3, 2017 File = D:\P466\hnd02-1.p466.a17.docm 1 http://faculty.washington.edu/jmiyamot/p466/p466-set.htm Psych 466: Judgment and Decision Making Autumn 2017

More information

Introduction to Research Methods

Introduction to Research Methods Introduction to Research Methods 8-10% of the AP Exam Psychology is an empirical discipline. Psychologists develop knowledge by doing research. Research provides guidance for psychologists who develop

More information

COMMITMENT &SOLUTIONS UNPARALLELED. Assessing Human Visual Inspection for Acceptance Testing: An Attribute Agreement Analysis Case Study

COMMITMENT &SOLUTIONS UNPARALLELED. Assessing Human Visual Inspection for Acceptance Testing: An Attribute Agreement Analysis Case Study DATAWorks 2018 - March 21, 2018 Assessing Human Visual Inspection for Acceptance Testing: An Attribute Agreement Analysis Case Study Christopher Drake Lead Statistician, Small Caliber Munitions QE&SA Statistical

More information

Classifying Substance Abuse among Young Teens

Classifying Substance Abuse among Young Teens Classifying Substance Abuse among Young Teens Dylan Rhodes, Sunet: dylanr December 14, 2012 Abstract This project attempts to use machine learning to classify substance abuse among young teens. It makes

More information

mirroru: Scaffolding Emotional Reflection via In-Situ Assessment and Interactive Feedback

mirroru: Scaffolding Emotional Reflection via In-Situ Assessment and Interactive Feedback mirroru: Scaffolding Emotional Reflection via In-Situ Assessment and Interactive Feedback Liuping Wang 1, 3 wangliuping17@mails.ucas.ac.cn Xiangmin Fan 1 xiangmin@iscas.ac.cn Feng Tian 1 tianfeng@iscas.ac.cn

More information

Client Care Counseling Critique Assignment Osteoporosis

Client Care Counseling Critique Assignment Osteoporosis Client Care Counseling Critique Assignment Osteoporosis 1. Describe the counselling approach or aspects of different approaches used by the counsellor. Would a different approach have been more appropriate

More information

WRITTEN ASSIGNMENT 1 (8%)

WRITTEN ASSIGNMENT 1 (8%) WRITTEN ASSIGNMENT 1 (8%) The purpose of this first written assignment is to give you practice at thinking scientifically about psychological issues. It s highly likely you have come across claims and

More information

Learning Process. Auditory Training for Speech and Language Development. Auditory Training. Auditory Perceptual Abilities.

Learning Process. Auditory Training for Speech and Language Development. Auditory Training. Auditory Perceptual Abilities. Learning Process Auditory Training for Speech and Language Development Introduction Demonstration Perception Imitation 1 2 Auditory Training Methods designed for improving auditory speech-perception Perception

More information

The Ordinal Nature of Emotions. Georgios N. Yannakakis, Roddy Cowie and Carlos Busso

The Ordinal Nature of Emotions. Georgios N. Yannakakis, Roddy Cowie and Carlos Busso The Ordinal Nature of Emotions Georgios N. Yannakakis, Roddy Cowie and Carlos Busso The story It seems that a rank-based FeelTrace yields higher inter-rater agreement Indeed, FeelTrace should actually

More information

How to Become Better at Lie Detection. Aldert Vrij University of Portsmouth Psychology Department

How to Become Better at Lie Detection. Aldert Vrij University of Portsmouth Psychology Department How to Become Better at Lie Detection Aldert Vrij University of Portsmouth Psychology Department Email: aldert.vrij@port.ac.uk Key points of my talk People are poor human lie detectors Cues to deception

More information

Secret Intelligence Service Room No. 15. Telling Lies: The Irrepressible Truth?

Secret Intelligence Service Room No. 15. Telling Lies: The Irrepressible Truth? Secret Intelligence Service Room No. 15 Telling Lies: The Irrepressible Truth? Emma J. Williams Lewis A. Bott John Patrick Michael B. Lewis 03 04 2013 Telling a lie takes longer than telling the truth

More information

PERSON PERCEPTION AND INTERPERSONAL ATTRACTION

PERSON PERCEPTION AND INTERPERSONAL ATTRACTION Person Perception and Interpersonal Attraction MODULE-V 22 PERSON PERCEPTION AND INTERPERSONAL ATTRACTION We have already noted, achieving a sense of self is an important achievement. A neonate may not

More information

STAT 200. Guided Exercise 4

STAT 200. Guided Exercise 4 STAT 200 Guided Exercise 4 1. Let s Revisit this Problem. Fill in the table again. Diagnostic tests are not infallible. We often express a fale positive and a false negative with any test. There are further

More information

Unsupervised Measurement of Translation Quality Using Multi-engine, Bi-directional Translation

Unsupervised Measurement of Translation Quality Using Multi-engine, Bi-directional Translation Unsupervised Measurement of Translation Quality Using Multi-engine, Bi-directional Translation Menno van Zaanen and Simon Zwarts Division of Information and Communication Sciences Department of Computing

More information

These methods have been explained in partial or complex ways in the original Mind Reading book and our rare Risk Assessment book.

These methods have been explained in partial or complex ways in the original Mind Reading book and our rare Risk Assessment book. These methods have been explained in partial or complex ways in the original Mind Reading book and our rare Risk Assessment book. In this lesson we will teach you a few of these principles and applications,

More information

Modeling State Space Search Technique for a Real World Adversarial Problem Solving

Modeling State Space Search Technique for a Real World Adversarial Problem Solving Modeling State Space Search Technique for a Real World Adversarial Problem Solving Kester O. OMOREGIE Computer Science Department, Auchi Polytechnic, Auchi, NIGERIA Stella C. CHIEMEKE Computer Science

More information

Assigning B cell Maturity in Pediatric Leukemia Gabi Fragiadakis 1, Jamie Irvine 2 1 Microbiology and Immunology, 2 Computer Science

Assigning B cell Maturity in Pediatric Leukemia Gabi Fragiadakis 1, Jamie Irvine 2 1 Microbiology and Immunology, 2 Computer Science Assigning B cell Maturity in Pediatric Leukemia Gabi Fragiadakis 1, Jamie Irvine 2 1 Microbiology and Immunology, 2 Computer Science Abstract One method for analyzing pediatric B cell leukemia is to categorize

More information

1. Stating the purpose of the speech is the first step in creating a speech.

1. Stating the purpose of the speech is the first step in creating a speech. 2 From A to Z: Overview of a Speech True/False Questions 1. Stating the purpose of the speech is the first step in creating a speech. 2. Audience analysis involves making random guesses about how the audience

More information

What Yelp Fake Review Filter Might Be Doing?

What Yelp Fake Review Filter Might Be Doing? Proceedings of the Seventh International AAAI Conference on Weblogs and Social Media What Yelp Fake Review Filter Might Be Doing? Arjun Mukherjee Vivek Venkataraman Bing Liu Natalie Glance University of

More information

Edge Level C Unit 4 Cluster 1 Face Facts: The Science of Facial Expressions

Edge Level C Unit 4 Cluster 1 Face Facts: The Science of Facial Expressions Edge Level C Unit 4 Cluster 1 Face Facts: The Science of Facial Expressions 1. Which group has been taught to read the clues in facial expressions? A. firefighters B. judges C. DEA agents D. border patrol

More information

Distributed by: Chart Your Course International Inc DISC - The Universal Language of Observable Behavior 1

Distributed by: Chart Your Course International Inc DISC - The Universal Language of Observable Behavior 1 D.I.S.C. The Universal Language of Observable Behavior Distributed by: Chart Your Course International Inc. www.chartcourse.com 800-821-2487 DISC - The Universal Language of Observable Behavior 1 DISC

More information

Education. Patient. Century. in the21 st. By Robert Braile, DC, FICA

Education. Patient. Century. in the21 st. By Robert Braile, DC, FICA Patient Education 21 st in the21 st Century By Robert Braile, DC, FICA Thealthcare marketplace. We also here are a few things we need to recognize relative to how chiropractic is perceived in the need

More information

Step 2 Challenging negative thoughts "Weeding"

Step 2 Challenging negative thoughts Weeding Managing Automatic Negative Thoughts (ANTs) Step 1 Identifying negative thoughts "ANTs" Step 2 Challenging negative thoughts "Weeding" Step 3 Planting positive thoughts 'Potting" Step1 Identifying Your

More information