Elamur Research⌕
☰
home / papers / 2026 / Profiling Psychopathic Behavior Using Machine Learning

Profiling Psychopathic Behavior Using Machine Learning

Paper · 2026 · Avi Treistman, Tehilla David, Sivan Levi et al.

Summary

The authors ask whether speech transcripts can reveal psychopathic traits, offering a screening aid that does not rely on clinical interviews alone, which are hard to scale and vulnerable to manipulation. They built an English corpus that contrasts interviews with confirmed criminal psychopaths, taken from YouTube, against a control group drawn from NPR interviews. Both sets were cleaned under shared annotation rules to minimise stylistic differences between sources. Instead of using the full vocabulary, they started from a hand-picked list of psychologically meaningful words. They expanded it with DyVaT, an embedding-based method that discards noisy high-variance dimensions, producing a compact lexicon. Texts were split into fixed-size chunks, lemmatized, vectorised with TF-IDF, and the minority class was enlarged through synonym substitution. Of three classical classifiers, a Linear SVM performed best, reaching roughly 81% accuracy. Inspecting feature weights exposed interpretable lexical contrasts between the groups. The work matters as an interpretable baseline for a domain where data is scarce and sensitive. It also flags ethical, bias, and privacy concerns for forensic and clinical deployment.

Key points

  • Data: The psychopathic class comes from about 77 YouTube interviews with legally and psychologically confirmed criminal psychopaths. Controls come from NPR interviews matched in conversational style. Transcripts were segmented into standardized 15-sentence chunks.
  • Features: Using Treistman et al.'s (2022) DyVaT (Dynamic Variance Thresholding) method, a manually built seed list was grown into a lexicon of about 1,357 features. DyVaT uses the kneedle method to choose a cut-off and keeps the low-variance embedding dimensions.
  • Models and results: Logistic Regression, Random Forest and Linear SVM were compared. The Linear SVM won with an F1 of 0.7957. Its accuracy is reported as 0.8081 in the abstract and introduction but as 0.8031 in the concluding summary.
  • Linguistic findings: Psychopaths' speech leaned more on basic-need words (money, food, sex) and on violence, dominance and authority vocabulary. Controls used more words tied to social bonds, norms and morality.
  • Limitations: The dataset is small and English-only, uses text without paralinguistic cues, comes from heterogeneous and partly censored sources, and rests on a contested definition of psychopathy. Planned extensions include LLMs, transformers, multilingual data and multimodal (audio/video) input.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Jerusalem College of TechnologyJerusalem · Israel13 works · 147 citations

 Fields, topics and concepts

📚 Shelves 7The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksPsychology · 1,552 worksSocial Psychology · 306 worksClinical Psychology · 20 works

Cite

Cite as Avi Treistman, Tehilla David, Sivan Levi and Dror Mughaz. 2026. Profiling Psychopathic Behavior Using Machine Learning. Proceedings of the Sixth Resources and ProcessIng of linguistic, para-linguistic and extra-linguistic Data from people with various forms of cognitive/psychiatric/developmental impairments in cooperation with the MENTAL.ai consortium.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.63317/2rto98ehmp4d

@inproceedings{treistman2026profiling,
  title = {{Profiling Psychopathic Behavior Using Machine Learning}},
  author = {Avi Treistman, Tehilla David, Sivan Levi and Dror Mughaz},
  booktitle = {Proceedings of the Sixth Resources and ProcessIng of linguistic, para-linguistic and extra-linguistic Data from people with various forms of cognitive/psychiatric/developmental impairments in cooperation with the MENTAL.ai consortium},
  year = {2026},
  doi = {10.63317/2rto98ehmp4d},
  url = {https://doi.org/10.63317/2rto98ehmp4d}
}

References

43 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Jonathan Adkins, Ali Al Bataineh, and Anthos Khanal. 2025. A psycholinguistic nlp framework for forensic text analysis of deception and emotion. Frontiers in Artificial Intelligence, 8:1669542. find
  2. Saqib Alam and Nianmin Yao. 2019. The impact of preprocessing steps on the accuracy of machine learning algorithms in sentiment analysis. Computational and Mathematical Organization Theory, 25(3):319–335. find
  3. Hesham Allam, Lisa Makubvure, Benjamin Gyamfi, Kwadwo Nyarko Graham, and Kehinde Akinwol-ere. 2025. Text classification: How machine learning is revolutionizing text categorization. Information, 16(2):130. find
  4. Yehia Ibrahim Alzoubi, Ahmet E Topcu, and Ahmed Enis Erkaya. 2023. Machine learning-based text classification comparison: Turkish language context. Applied Sciences, 13(16):9428. find
  5. Avanti Bhandarkar, Ronald Wilson, Anushka Swarup, Gregory Webster, and Damon Woodard. 2025. Psytex: A knowledge-guided approach to refining text for psychological analysis. In Proceedings of the 5th International Conference on Natural Language Processing for Digital Humanities, pages 151–178. find
  6. Evan Bose, Chaitanya Anil Kumar, and N Meenakshi. 2025. Ai-driven psychological profiling on social media: Mechanisms, ethical breaches, and regulatory challenges in data inference. recent trends in social studies. 2025; 2 (1): 1–7p. AI-Driven Psychological Profiling on Social Media Bose et al. STM… find
  7. A Bouguettaya, EM Stuart, and E Aboujaoude. Racial bias in ai-mediated psychiatric diagnosis and treatment: a qualitative comparison of four large language models. npj digit. med. 8, 332 (2025). find
  8. Alexis Carrillo, Simon Friedrich Roske, Rebeca Ianov-Vitanov, Enrico Perinelli, Alessandro Gre-cucci, and Massimo Stella. 2025. Textual forma mentis networks bridge language structure, emotional content and psychopathology levels in adolescents. arXiv preprint arXiv:2505.06387. — [arXiv:2505.06387](https://arxiv.org/abs/2505.06387) find
  9. Vivek Chavan, Arsen Cenaj, Shuyuan Shen, Ariane Bar, Srishti Binwani, Tommaso Del Becaro, Marius Funk, Lynn Greschner, Roberto Hung, Stina Klein, et al. 2025. Feeling machines: Ethics, culture, and the rise of emotional ai. arXiv preprint arXiv:2506.12437. — [arXiv:2506.12437](https://arxiv.org/abs/2506.12437) find
  10. Jiaao Chen, Derek Tam, Colin Raffel, Mohit Bansal, and Diyi Yang. 2023. An empirical survey of data augmentation for limited data learning in nlp. Transactions of the Association for Computational Linguistics, 11:191–211.
  11. Mamata Das, PJA Alphonse, et al. 2023. A comparative study on tf-idf feature weighting method and its analysis using unstructured dataset. arXiv preprint arXiv:2308.04037. — [arXiv:2308.04037](https://arxiv.org/abs/2308.04037) find
  12. Krishna Kant Dixit, Sumit Pundir, Anurag Shrivastava, C Praveen Kumar, Arun Pratap Srivastava, and Pankaj Singh. 2023. Analyzing textual data for mental health assessment: Natural language processing for depression and anxiety. In 2023 10th IEEE Uttar Pradesh Section International Conference on Elec… find
  13. Andrea Gasparetto, Matteo Marcuzzo, Alessandro Zangari, and Andrea Albarelli. 2022. A survey on text classification algorithms: From text to predictions. Information, 13(2):83. find
  14. Barbara Gawda. 2022. The differentiation of narrative styles in individuals with high psychopathic deviate. Journal of Psycholinguistic Research, 51(1):75–92. find
  15. Hamideh Ghanadian, Isar Nejadgholi, and Hussein Al Osman. 2025. Improving suicidal ideation detection in social media posts: Topic modeling and synthetic data augmentation approach. JMIR Formative Research, 9:e63272. find
  16. Yuting Guo, Anthony Ovadje, Mohammed Ali Al-Garadi, and Abeed Sarker. 2024. Evaluating large language models for health-related text classification tasks with public social media data. Journal of the American Medical Informatics Association, 31(10):2181–2189. find
  17. J. T. Hancock, M. Woodworth, and R. Boochever. 2018. Psychopaths online: The linguistic traces of psychopathy in email, text messaging and facebook. Media and Communication, 6(3):83–92. find
  18. R. D. Hare. 2020. The pcl-r assessment of psychopathy. In The Wiley International Handbook on Psychopathic Disorders and the Law, pages 63–106. Wiley. find
  19. Jashanjot Kaur and P Kaur Buttar. 2018. A systematic review on stopword removal algorithms. International Journal on Future Revolution in Computer Science & Communication Engineering, 4(4):207–210. find
  20. Xingxuan Li, Yutong Li, Linlin Liu, Lidong Bing, and Shafiq Joty. 2022. Is gpt-3 a psychopath? evaluating large language models from a psychological perspective. arXiv preprint arXiv:2212.10529, 10. — [arXiv:2212.10529](https://arxiv.org/abs/2212.10529) find
  21. Xingxuan Li, Yutong Li, Lin Qiu, Shafiq Joty, and Lidong Bing. 2024. Evaluating psychological safety of large language models. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 1826–1843. with emotion representation. Frontiers in Psychology, 16:1612769. find
  22. Leberecht Maxim, Nedderhoff Andre, Zitzmann Steffen, and Hecht Martin. 2025. Comparing machine learning methods for predicting dark triad personality traits using social media text data. Journal of Research in Personality, page 104690. find
  23. Brent Mittelstadt. 2019. Principles alone cannot guarantee ethical ai. Nature machine intelligence, 1(11):501–507. find
  24. Dena F Mujtaba and Nihar R Mahapatra. 2025. Behind the screens: Uncovering bias in ai-driven video interview assessments using counterfactuals. arXiv preprint arXiv:2505.12114. — [arXiv:2505.12114](https://arxiv.org/abs/2505.12114) find
  25. Anam Naz, Hikmat Ullah Khan, Amal Bukhari, Bader Alshemaimri, Ali Daud, and Muhammad Ramzan. 2025. Machine and deep learning for personality traits detection: a comprehensive survey and open research challenges. Artificial Intelligence Review, 58(8):239. find
  26. NPR. Npr media dialog transcripts. Kaggle Dataset. Accessed: September 22, 2025. find
  27. Dhruv Patel and Anju Johnson. 2025. Detecting narcissistic personality disorder (npd): A hybrid regex and nlp based ai approach with phase-aware classification. IEEE Access. find
  28. Rilo Chandra Pradana and Derwin Suhartono. 2024. Synonym replacement augmentation for handling data imbalance in personality classification. In 2024 8th International Conference on Information Technology, Information Systems and Electrical Engineering (ICITISEE), pages 1– 6. IEEE. find
  29. Md Ashrafur Rahman, Evangelos Victoros, Rob Davis, Tariq Duaa, Yeasna Shanjana, and Md Rabiul Islam. 2025. Use of artificial intelligence in mental healthcare, health psychology, and related research: A narrative review to address challenges and opportunities. Health Science Reports, 8(12):e71595. find
  30. Jingfang Liu, Peng Ding, and Jie Chen. 2025. Depglm: Depression degree recognition on social media based on large language models. Digital Health, 11:20552076251408281. find
  31. María E Raygoza-L, Jesús Heriberto Orduño-Osuna, Roxana Jimenez-Sanchez, and Fabian N Murrieta-Rico. 2025. Innovative artificial intelligence approaches for identifying and managing dsm cluster b personality disorders in mental health: A case study on the dark triad. In Exploring Psychology, Social… find
  32. Huimin Mao and Qing Han. 2025. Enhancing textgcn for depression detection on social media Minhah Saleem and Jihie Kim. 2024. Intent aware data augmentation by leveraging generative ai for stress detection in social media texts. PeerJ Computer Science, 10:e2156. find
  33. Bosubabu Sambana, Kondreddygari Archana, Suram Indhra Sena Reddy, Shaik Meethaigar Jameer Basha, and Shaik Karishma. 2025. Data augmentation for cognitive behavioral therapy: Leveraging ernie language models using artificial intelligence. In 2025 3rd International Conference on Intelligent Data Comm… find
  34. Sumona Yeasmin, Nazia Nowshin, and Tasnia Afrin Chowdhury. 2024. Identifying human dark triad from text data through machine learning models. International Journal of Research and Innovation in Applied Science, 9(6):89–104. find
  35. Jianlong Zhou and Fang Chen. 2023. Ai ethics: From principles to practice. Ai & Society, 38(6):2693–2703. find
  36. Lingzhi Shen, Yunfei Long, Xiaohao Cai, Guanming Chen, Yuhan Wang, Imran Razzak, and Shoaib Jameel. 2025. Ll4g: Self-supervised dynamic optimization for graph-based personality detection. arXiv preprint arXiv:2504.02146. — [arXiv:2504.02146](https://arxiv.org/abs/2504.02146) find
  37. Zhivar Sourati, Meltem Ozcan, Colin McDaniel, Alireza Ziabari, Nuan Wen, Ala Tak, Fred Morstatter, and Morteza Dehghani. 2024. Secret keepers: The impact of llms on linguistic markers of personal traits. arXiv preprint arXiv:2404.00267. — [arXiv:2404.00267](https://arxiv.org/abs/2404.00267) find
  38. Tom Sühr, Florian E Dorner, Samira Samadi, and Augustin Kelava. 2025. Challenging the validity of personality tests for large language models. In Proceedings of the 2025 Equity and Access in Algorithms, Mechanisms, and Optimization, pages 74–81. find
  39. Daniela Teodorescu, Tiffany Cheng, Alona Fyshe, and Saif Mohammad. 2023. Language and mental health: Measures of emotion dynamics from text as linguistic biosocial markers. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 3117–3133. find
  40. Avraham Treistman, Dror Mughaz, Ariel Stulman, and Amit Dvir. 2022. Word embedding dimensionality reduction using dynamic variance thresholding (DyVaT). Expert Systems with Applications, 208:118157. find
  41. Yuli Vasiliev. 2020. Natural Language Processing with Python and spaCy. Packt Publishing. Accessed: December 6, 2025. find
  42. Michael Veale and Borgesius Frederik Zuiderveen. 2021. Demystifying the draft eu artificial intelligence act. Computer Law Review International, 22(4):97–112. find
  43. G Venkateshwarlu, S Akhila, V Kavyasree, S Vishnu, and VS Prasad. 2024. Enhanced text classification using random forest: Comparative analysis and insights on performance and efficiency. Int. J. Comput. Eng. Res. Trends, 11:1–8. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/treistman-2026-profiling-psychopathic-behavior-machine-learning/