Combining psycho-linguistic, content-based and chat-based features to detect predation in chatrooms
Open accessAbstract
Abstract: The Digital Age has brought great benefits for the human race but also some draw-backs. Nowadays, people from opposite corners of the World can communicate online via instant messaging services. Unfortunately, this has introduced new kinds of crime. Sexual predators have adapted their predatory strategies to these platforms and, usually, the target victims are kids. The authorities cannot manually track all threats because massive amounts of online conversations take place in a daily basis. Automatic methods for alerting about these crimes need to be de-signed. This is the main motivation of this paper, where we present a Machine Learning approach to identify suspicious subjects in chat-rooms. We propose novel types of features for represen-ting the chatters and we evaluate different classifiers against the largest benchmark available. This empirical validation shows that our approach is promising for the identification of predatory behaviour. Furthermore, we carefully analyse the characteristics of the learnt classifiers. This preliminary analysis is a first step towards profiling the behaviour of the sexual predators when chatting on the Internet.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Parpar et al. categorized 80 types of LIWC features that were used for predatory detection in chatrooms.
“Parpar et al. [70] have categorized 80 types of LIWC features that were used for predatory detection in chatrooms.”
From Online grooming detection: A comprehensive survey of child exploitation in chat logs (2023) · cites this paper as [70]
The Introduction (p. 214) says the authors used psycho-linguistic evidence to define features for classifiers detecting sexual predation in chatrooms. The available passages do not identify those features as LIWC categories or give a count of 80; searches for “LIWC” and “80 categories” found no supporting passage. (LLM, conf=0.78)
Experimental — accuracy not yet measuredThe activity of the author and the time of the chats, such as late at night, could be used as a feature for detecting predatory chats.
“Parapar et al. [70] pointed out that the activity of the author and the time of the chats (e.g., if the chats happened late at night) could also be used as a feature for detecting predatory chats.”
From Online grooming detection: A comprehensive survey of child exploitation in chat logs (2023) · cites this paper as [70]
The Introduction (p. 214) says the classifiers include features based on chatters’ global activity, supporting the activity part of the claim. The passages retrieved for chat timing and nighttime activity do not say that the time of a chat, including late at night, was proposed or used as a feature. (LLM, conf=0.82)
Experimental — accuracy not yet measuredThe deceptive trust phase and language pattern include effective use of loving, time, and location words.
“Their results indicated that the deceptive trust phase and language pattern include effective use of loving, time, and location words.”
From Online grooming detection: A comprehensive survey of child exploitation in chat logs (2023) · cites this paper as [70]
The Introduction (pp. 214–215) describes sexual predation as deceptive and says word-use features can help identify predators. The passages available from our copy do not establish a distinct deceptive-trust phase or report that loving, time, and location words characterize it; searches for those terms found no such result. (LLM, conf=0.72)
Experimental — accuracy not yet measuredEmotional expression could be a good indicator of the deceptive language of a predator.
“The authors also showed that emotional expression could be a good indicator of the deceptive language of a predator.”
From Online grooming detection: A comprehensive survey of child exploitation in chat logs (2023) · cites this paper as [70]
The time of chat was observed closer to midnight.
“Other features such as time of chat were observed closer to midnight, and predators' linguistic profiles showed that they mostly use first-person pronouns [70].”
From Online grooming detection: A comprehensive survey of child exploitation in chat logs (2023) · cites this paper as [70]
The Introduction (page not provided) says the authors analyse chatters’ global activity, but the passages retrieved do not report when chats occurred. Searches for midnight, night, time of day, and chat hours found no observation that chats occurred closer to midnight. (LLM, conf=0.72)
Experimental — accuracy not yet measured
Content-based features indicated characteristics such as how active, anxious, or intense an author was.
“Parapar et al. [70] also noted that the content-based features indicated characteristics such as how active, anxious, or intense an author was.”
From Online grooming detection: A comprehensive survey of child exploitation in chat logs (2023) · cites this paper as [70]
The available Introduction (p. 214) says the authors compute content-based features from chatters’ text, but it does not say those features indicate how active, anxious, or intense an author is. Searches for those terms returned no relevant passage; the accessible text appears to stop in the Introduction, so the potentially decisive feature descriptions may be missing. (LLM, conf=0.72)
Experimental — accuracy not yet measuredPredators' linguistic profiles showed that they mostly use first-person pronouns.
“Other features such as time of chat were observed closer to midnight, and predators' linguistic profiles showed that they mostly use first-person pronouns [70].”
From Online grooming detection: A comprehensive survey of child exploitation in chat logs (2023) · cites this paper as [70]
The Introduction says pronouns can provide clues about deception and are used as features in the authors’ predator-detection classifiers. It does not say predators mostly use first-person pronouns. Searches for first-person pronouns and predator linguistic profiles found no passage establishing that specific pattern. (LLM, conf=0.68)
Experimental — accuracy not yet measuredPredators engage primarily in 1-on-1 conversations and less in conversations involving multiple persons.
“Their study reported that predators engage primarily in 1-on-1 conversations and less in conversations involving multiple persons.”
From Online grooming detection: A comprehensive survey of child exploitation in chat logs (2023) · cites this paper as [70]
The available Introduction passages (p. 214) say the study analyzes predators’ chat behavior, but do not compare one-on-one with multiperson conversations. Searches for “1-on-1,” “multi-party conversations,” and the number of participants returned no passage reporting that comparison. (LLM, conf=0.72)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Text categorization with Support Vector Machines: Learning with many relevant features
Joachims, T. (1998) · Lecture notes in computer science · doi:10.1007/bfb0026683 · Cited by 8,134 · see citing works
Machine learning in automated text categorization
Sebastiani, F. (2002) · ACM Computing Surveys · doi:10.1145/505282.505283 · Cited by 7,927 · see citing works
LIBLINEAR: A Library for Large Linear Classification
Fan, R.-E., Chang, K., Hsieh, C., Xiang-rui, W., & Lin, C. (2008) · doi:10.5555/1390681.1442794 · Cited by 6,680 · see citing works
A Practical Guide to Support Vector Classication
Hsu, C., Chang, C.-C., & Lin, C. (2008) · Cited by 5,070 · see citing works
Psychological Aspects of Natural Language Use: Our Words, Our Selves
Pennebaker, J. W., Mehl, M. R., & Niederhoffer, K. G. (2003) · Annual Review of Psychology · doi:10.1146/annurev.psych.54.101601.145041 · Cited by 2,624 · see citing works
Linguistic styles: language use as an individual difference
Pennebaker, J. W., & King, L. A. (1999) · Journal of Personality and Social Psychology · doi:10.1037/0022-3514.77.6.1296 · Cited by 1,961 · see citing works
The Development and Psychometric Properties of LIWC2015
Pennebaker, J. W., Boyd, R. L., Jordan, K., & Blackburn, K. (2015) · University of Texas at Austin · doi:10.15781/t25p41 · Cited by 1,674 · see citing works
Inductive learning algorithms and representations for text categorization
Dumais, S., Platt, J. C., Heckerman, D. E., & Sahami, M. (1998) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/288627.288651 · Cited by 1,463 · see citing works
Lying Words: Predicting Deception from Linguistic Styles
Newman, M. L., Pennebaker, J. W., Berry, D. S., & Richards, J. M. (2003) · Personality and Social Psychology Bulletin · doi:10.1177/0146167203029005010 · Cited by 1,441 · see citing works
Finding Deceptive Opinion Spam by Any Stretch of the Imagination
Ott, M., Choi, Y., Cardie, C., & Hancock, J. T. (2011) · arXiv · doi:10.48550/arxiv.1107.4557 · Cited by 1,008 · see citing works
Detecting Lies and Deceit: The Psychology of Lying and the Implications for Professional Practice
Vrij, A. (2000) · Medical Entomology and Zoology · Cited by 627 · see citing works
On Lying and Being Lied To: A Linguistic Analysis of Deception in Computer-Mediated Communication
Hancock, J. T., Curry, L. E., Goorha, S., & Woodworth, M. (2007) · Discourse Processes · doi:10.1080/01638530701739181 · Cited by 523 · see citing works
Androutsopoulos, I., Koutsias, J., Chandrinos, K. V., & Spyropoulos, C. D. (2000) · Proceedings of the 23rd annual international ACM SIGIR conference on Research and development in information retrieval · doi:10.1145/345508.345569 · Cited by 465 · see citing works
Mihalcea, R. F., & Strapparava, C. (2009) · doi:10.3115/1667583.1667679 · Cited by 341 · see citing works
Discriminative models for information retrieval
Nallapati, R. M. (2004) · International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) · doi:10.1145/1008992.1009006 · Cited by 288 · see citing works
Feature Ranking Using Linear SVM
Chang, Y.-W., & Lin, C. (2008) · Cited by 246 · see citing works
Author gender identification from text
Cheng, N., Chandramouli, R., & Subbalakshmi, K. P. (2011) · Digital Investigation · doi:10.1016/j.diin.2011.04.002 · Cited by 224 · see citing works
Investigative Data Mining for Security and Criminal Detection
Mena, J. (2002) · Cited by 187 · see citing works
Matsumoto, Y., & Mihalcea, R. F. (2011) · Cited by 172 · see citing works
Verbal behavior in everyday life
Weintraub, W. (1989) · Cited by 143 · see citing works
Machiavellianism and Self-Monitoring: As Different as “Me” and “You”
Ickes, W. J., Reidhead, S. M., & Patterson, M. L. (1986) · Social Cognition · doi:10.1521/soco.1986.4.1.58 · Cited by 131 · see citing works
Malesky, L. A. (2007) · Journal of Child Sexual Abuse · doi:10.1300/j070v16n02_02 · Cited by 128 · see citing works
Learning to Identify Internet Sexual Predation
McGhee, I., Bayzick, J., Kontostathis, A., Edwards, L., McBride, A., & Jakubowski, E. (2011) · International Journal of Electronic Commerce · doi:10.2753/jec1086-4415150305 · Cited by 106 · see citing works
Grammatical word class variation within the British National Corpus Sampler
Rayson, P., Wilson, A., & Leech, G. (2002) · BRILL eBooks · doi:10.1163/9789004334113_020 · Cited by 104 · see citing works
Overview of the International Sexual Predator Identification Competition at PAN-2012.
Inches, G., & Crestani, F. (2012) · Cited by 93 · see citing works
Design and Analysis of the Causation and Prediction Challenge
Guyon, I., Aliferis, C. C., Cooper, G. F., Pellet, J., Spirtes, P., & Statnikov, A. (2008) · Cited by 78 · see citing works
(1993) · BRILL eBooks · doi:10.1163/9789004653559_024 · Cited by 77 · see citing works
Interpreting the Intentions of Internet Predators: An Examination of Online Predatory Behavior
Marcum, C. D. (2007) · Journal of Child Sexual Abuse · doi:10.1300/j070v16n04_06 · Cited by 77 · see citing works
Feature Selection Using Linear Support Vector Machines
Brank, J., & Grobelnik, M. (2002) · Cited by 75 · see citing works
Reliability of human annotation of semantic roles in noisy text
Higgins, D. (2007) · IEEE International Conference on Semantic Computing (ICSC) · doi:10.1109/icsc.2007.102 · Cited by 74 · see citing works
EFFECTIVENESS OF SUPPORT VECTOR MACHINE FOR CRIME HOT-SPOTS PREDICTION
Kianmehr, K., & Alhajj, R. (2008) · Applied Artificial Intelligence · doi:10.1080/08839510802028405 · Cited by 41 · see citing works
(2012) · Cited by 28 · see citing works
A Learning-Based Approach for the Identification of Sexual Predators in Chat Logs.
Parapar, J., Losada, D. E., & Barreiro, Á. (2012) · CLEF (Online Working Notes/Labs/Workshop) · Cited by 27 · see citing works
Self-citationConversation Level Constraints on Pedophile Detection in Chat Rooms
Peersman, C., Vaassen, F., Asch, V. V., & Daelemans, W. M. P. (2012) · Bristol Research (University of Bristol) · Cited by 23 · see citing works
Identifying Sexual Predators by SVM Classification with Lexical and Behavioral Features.
Morris, C. W., & Hirst, G. (2012) · Cited by 23 · see citing works
Modelling Fixated Discourse in Chats with Cyberpedophiles
Bogdanova, D., Rosso, P., & Solorio, T. (2012) · Cited by 22 · see citing works
Gillam, L., & Vartapetiance, A. (2012) · Surrey Research Insight Open Access (The University of Surrey) · Cited by 20 · see citing works
EXPLORATION OF MASSIVE CRIME DATA SETS THROUGH DATA MINING TECHNIQUES
Lee, I., & Estivill‐Castro, V. (2011) · Applied Artificial Intelligence · doi:10.1080/08839514.2011.570153 · Cited by 19 · see citing works
AN OVERVIEW OF DATA MINING FOR COMBATING CRIME
Nissan, E. (2012) · Applied Artificial Intelligence · doi:10.1080/08839514.2012.713309 · Cited by 16 · see citing works
Villatoro-Tello, E., Juárez-González, A., Escalante, H. J., Montes-y-Gómez, M., & Villaseñor-Pineda, L. (2012) · Cited by 14 · see citing works
Extending Textual Models of Deception to Interrogation Settings
Skillicorn, D. B., & Lamb, C. (2013) · Linguistic Evidence in Security Law and Intelligence · doi:10.5195/lesli.2013.1 · Cited by 4 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
eRISK 2017: CLEF Lab on Early Risk Prediction on the Internet: Experimental Foundations
Losada, D. E., Crestani, F., & Parapar, J. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-65813-1_30 · Cited by 164 · see citing works
Self-citationMartínez-Torres, M. R., & Toral, S. (2019) · Tourism Management · doi:10.1016/j.tourman.2019.06.003 · Cited by 105 · see citing works
Cyber-aggression, Cyberbullying, and Cyber-grooming: A Survey and Research Challenges
Mladenović, M., Ošmjanski, V., & Stanković, S. V. (2022) · ACM Computing Surveys · doi:10.1145/3424246 · Cited by 79 · see citing works
Early detection of deception and aggressiveness using profile-based representations
Escalante, H. J., Villatoro-Tello, E., Garza, S. E., López-Monroy, A. P., Montes-y-Gómez, M., & Villaseñor-Pineda, L. (2017) · Expert Systems with Applications · doi:10.1016/j.eswa.2017.07.040 · Cited by 44 · see citing works
Ali, S., Razi, A., Kim, S., Alsoubai, A., Ling, C., Choudhury, M. D., Wiśniewski, P., & Stringhini, G. (2023) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3579608 · Cited by 33 · see citing works
Deception Detection and Opinion Spam
Rosso, P., & Cagnina, L. (2017) · Socio-affective computing · doi:10.1007/978-3-319-55394-8_8 · Cited by 16 · see citing works
Online Social Networks and Writing Styles–A Review of the Multidisciplinary Literature
Tai, K. Y., Dhaliwal, J., & Shariff, S. M. (2020) · IEEE Access · doi:10.1109/access.2020.2985916 · Cited by 10 · see citing works
Deception Detection for the Russian Language: Lexical and Syntactic Parameters
Борисовна, П. Д., Litvinova, T., & Litvinova, O. (2017) · doi:10.26615/978-954-452-038-0_001 · Cited by 8 · see citing works
Managing Emotion In The Workplace: An Empirical Study With Enterprise Instant Messaging
Ke, S., Tsai, C., & Chen, Y.-J. (2024) · Applied Artificial Intelligence · doi:10.1080/08839514.2023.2297518 · Cited by 7 · see citing works
Dissecting Genuine and Deceptive Kudos: The Case of Online Hotel Reviews
Banerjee, S., & Chua, A. Y. K. (2014) · International Journal of Advanced Computer Science and Applications · doi:10.14569/specialissue.2014.040304 · Cited by 5 · see citing works
A survey on Datamining in Cyber Bullying
Nalini, K., & Sheela, L. J. (2014) · Cited by 3 · see citing works
Enabling Public Security Text-Based Analytics: A Survey to Outline Research DirectionsPreprint
de Carvalho, V. D. H., dos Santos, R. J. R., Nepomuceno, T. C. C., & Poleto, T. (2024) · Preprints.org · doi:10.20944/preprints202403.0064.v1 · Cited by 2 · see citing works
Classification Model for Bullying Posts Detection
Nalini, K., & Jabasheela, D. L. (2019) · IntechOpen eBooks · doi:10.5772/intechopen.88633
Detecting an Autobiographical Criminal: Investigating Gender Differences in Staged Suicide Notes
Gill, K. G. (2019) · e-space (Manchester Metropolitan University)
Reducing Sexual Predation and Victimization Through Warnings and Awareness among High-Risk UsersPreprint
Takano, M., Nishiguchi, M., & Toriumi, F. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5173981
Reducing sexual predation and victimization through warnings and awareness among high-risk users
Takano, M., Nishiguchi, M., & Toriumi, F. (2025) · Journal of Computational Social Science · doi:10.1007/s42001-025-00399-3
- Authors
- Barreiro, A., Losada, D. E., & Parapar, J.
- Venue
- Verlag der Technischen Universität Graz
- Publisher
- Verlag der Technischen Universität Graz
- Published
- 2014
- OpenAlex
- W2293649141
- Cited by
- 16
- References
- 41
- Citations by year
- 2025: 2 · 2024: 2 · 2023: 1 · 2021: 1 · 2020: 1 · 2019: 3 · 2017: 4 · 2014: 2
- FWCI
- 0.29
- Type
- journal article
- Language
- en
- Open access
- green
- zenodo.org
- Licence
- other-oa
- Locations
- Zenodo (CERN European Organization for Nuclear Research) (submittedVersion, other-oa)
- TUGraz OPEN Library (Graz University of Technology)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Hate Speech and Cyberbullying Detection
- Physical Sciences › Computer Science › Artificial Intelligence › Authorship Attribution and Profiling
- Physical Sciences › Computer Science › Information Systems › Spam and Phishing Detection
- Keywords
- instant messaging, content-based features, child online safety
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
