### Bennett Kleinberg I am an Associate Professor in Behavioural Data Science at the [Department of Methodology and Statistics](https://www.tilburguniversity.edu/about/schools/socialsciences/organization/departments/methodology-statistics) at Tilburg University and an Honorary Associate Professor at the [Department of Security and Crime Science](http://www.ucl.ac.uk/jill-dando-institute) at University College London. Previously I've held positions with the [Dawes Centre for Future Crime](http://www.ucl.ac.uk/jill-dando-institute/dawes-future-crime) at University College London and the University of Amsterdam's [Department of Psychology](https://www.uva.nl/en/about-the-uva/organisation/faculties/faculty-of-social-and-behavioural-sciences/disciplines/psychology/psychology.html). I obtained my PhD form the Department of Psychology at the University of Amsterdam. --- - [Google Scholar](https://scholar.google.nl/citations?user=Ms2d7OsAAAAJ&hl=en) - [GitHub](https://github.com/ben-aaron188) - [Semantic Scholar](https://www.semanticscholar.org/author/Bennett-Kleinberg/6032930) --- #### Research My research interests revolve around behavioural inference with two core questions: (1) How can we study and understand human behaviour through computational methods? and (2) How can we make inferences about the behaviour of computational models with methods from the behavioural sciences? Of particular interest is the intersection between the behavioural (and social) sciences and natural language processing. Questions I'm currently interested in and working on: - How can we use text data to study human behaviour and cognition? - How can we formally measure the processes through which a large language model produces its outputs? - How (if at all!) can we use observational online data to _measure_ and _predict_ behaviour? And how can we (as online users) prevent this? - Which assumptions do we (need to) make when working with text data and are these valid assumptions? - What are the methodological pitfalls of either discipline and how can we fix this? - How can tools/methods of either discipline help the other? - How can we use NLP to ask better questions for deception detection? Other research interests are about applying computational techniques to questions around crime, deception, threat assessment (see the research areas of my PhD students below). I'm also one of the inventors of [Textwash](https://github.com/maximilianmozes/textwash) - an automated, learning-based text anonymisation software and [SAGE Ocean concept grant winner](https://uk.sagepub.com/en-gb/eur/press/sage-ocean-announces-text-wash-as-2019-concept-grant-winner). Some [(more info here)](https://ocean.sagepub.com/concept-grants) and [(interviews on YouTube)](https://www.youtube.com/watch?v=T9pRRn2DrMY). --- #### Research lab We're running a research lab that consists of core members (current and former PhD students) and thesis/internship/traineeship students. More info on this group via this [link](https://bkleinberg.net/whatever_research/) I have the privilege of supervising these amazing young doctoral researchers for their PhD: - Sanne Peereboom, _Assessing the artificial mind through the marriage of natural language processing and psychometrics_, 2023-2027 (Bennett Kleinberg/Inga Schwabe) - Weng Lam Ao, _Understanding decision-making in transport behaviour through social media data_, 2023-2027, joint PhD with University of Modena and Reggio Emilia, (Bennett Kleinberg/Francisco Bahamonde Birke) - Biniyam Gebeyehu, _Developing automated infection risk stratification with NLP from electronic health records_, 2021-2025 (Esther de Vries/Bennett Kleinberg/Katrijn van Deun) - Rasoul Norouzi Nikjeh, _Text-mining methods for theory development in social science research_, 2023-2027 (Caspar van Lissa/Bennett Kleinberg/Jeroen Vermunt) - Daniel Hammocks [(www)](https://danielhammocks.uk/), _Information prioritisation for horizon scanning using data science techniques_, 2019-2026 (Kate Bowers/Bennett Kleinberg) **PhD alumni:** - Dr Isabelle van der Vegt [(www)](http://isabellevdv.net/) - Linguistic Threat Assessment: Understanding Targeted Violence through Computational Linguistics (supervisors: Paul Gill/Bennett Kleinberg) - completed in Spring 2021 (now Assistant Prof at Utrecht University) - Dr Felix Soldner [(www)](http://fsoldner.net/) - Detecting and mitigating online consumer fraud (supervisors: Shane Johnson/Bennett Kleinberg) - completed in Summer 2023 (now Postdoc at GESIS, Cologne) - Maximilian Mozes [(www)](http://mmozes.net/) - Adversarial machine learning in natural language processing (supervisors: Lewis Griffin/Bennett Kleinberg) - viva in Fall 2023 - Arianna Trozze [(www)](https://ariannatrozze.github.io/) - The evidential chain in cryptocurrency crime (supervisors: Toby Davies/Bennett Kleinberg) - viva in Fall 2023 --- #### Teaching - 2023/2024 - Data Acquisition Methods (BSc in Data Science), Tilburg University + Eindhoven University of Technology - Data Science Challenge 2 (BSc in Data Science), Tilburg University + Eindhoven University of Technology - Machine Behaviour, TiU BSc honours programme course, Tilburg University - 2022/2023 - Introduction to Statistics (for the BSc programmes in Psychology, Sociology and Human Resource Studies, English + Dutch version - 1400 students), Tilburg University - Data Science Challenge 2 (BSc in Data Science - 150 students), Tilburg University + Eindhoven University of Technology - 2021/2022 - Introduction to Statistics (for the BSc programmes in Psychology, Sociology and Human Resource Studies, English + Dutch version - 1300 students), Tilburg University - Data Science challenge 2 (BSc in Data Science - 140 students), Tilburg University + Eindhoven University of Technology - Crime Science (interdisciplinary BSc elective - 100 students), University of Amsterdam [(study guide)](https://www.uva.nl/en/programmes/open-programmes-iis/crime-science/crime-science.html?cb) - 2020/2021 - Introduction to Statistics (for the BSc programmes in Psychology, Sociology and Human Resource Studies, English + Dutch version - 1300 students), Tilburg University - Data Science Challenge 1 (BSc in Data Science - 100 students), Tilburg University + Eindhoven University of Technology (received award for "Excellent Teaching Evaluation") - 2019/2020 - SECU0050, Data Science for Crime Scientists (BSc - 25 students), UCL, [Material on GitHub](https://github.com/ben-aaron188/UCL_SECU0050) - SECU0057, Applied Data Science (MSc - 20 students), UCL, [Material on GitHub](https://github.com/ben-aaron188/UCL_SECU0057) - 2018/2019 - SECU0013, Probability, Statistics & Modelling 2 (BSc - 25 students), UCL, [Material on GitHub](https://github.com/ben-aaron188/ucl_psm2_20182019) - SECU0050, Data Science for Crime Scientists (BSc - 20 students), UCL, [Material on GitHub](https://github.com/ben-aaron188/ucl_aca_20182019) - Crime Science (BSc honours module - 25 students), University of Amsterdam [(study guide)](https://www.uva.nl/en/programmes/honours-modules-iis/honoursmodule-crime-science/honoursmodule-crime-science.html?1549646204808), [(interview)](https://iis.uva.nl/en/content/news/2019/02/keuzevakcrimescience.html) - 2017/2018 - Crime Science (Open UvA module - 75 students), University of Amsterdam, [Module info](https://www.uva.nl/en/programmes/open-programmes-iis/crime-science/crime-science.html) - Crime Science (BSc honours module- 25 students), University of Amsterdam [(study guide)](http://www.studiegids.uva.nl/xmlpages/page/2017-2018-en/search-course/course/40888) - Deception and Lie Detection (MSc - 8 students), University of Amsterdam [(study guide)](http://www.studiegids.uva.nl/xmlpages/page/2017-2018-en/search-course/course/41168) - 2016/2017 - Deception and Lie Detection (MSc - 12 students), University of Amsterdam [(study guide)](http://www.studiegids.uva.nl/xmlpages/page/2016-2017-en/search-course/course/28244) - 2015/2016 - Deception and Lie Detection (MSc - 15 students), University of Amsterdam --- #### Media coverage **Written media** [European Scientist](https://www.europeanscientist.com/en/research/a-new-way-to-catch-liars/), [New York Post](https://nypost.com/2023/04/03/heres-how-to-tell-if-someone-is-lying-to-you/), [New Scientist](https://www.newscientist.com/article/2365049-is-someone-lying-to-you-a-lack-of-detail-may-give-them-away/), [Daily Mail](https://www.dailymail.co.uk/sciencetech/article-11885973/The-new-way-spot-LIAR-according-researchers.html), [TiU press release](https://www.tilburguniversity.edu/magazine/detect-mental-problems-text-data), [Ulyces (French)](https://www.ulyces.co/denis-hadzovic/drill-uk-des-roulements-de-caisses-claires-et-des-meurtres-au-couteau/), [UCL News II](https://www.ucl.ac.uk/news/2020/may/how-men-and-womens-anxieties-differ-over-covid-19), [Daily Mail](https://www.dailymail.co.uk/sciencetech/article-8316235/Men-worry-economy-COVID-19-women-concerned-loved-ones.html), [Disover Magazine](https://www.discovermagazine.com/the-sciences/the-curious-case-of-drill-rap-and-its-link-to-violent-crime), [SAGE Ocean press release](https://uk.sagepub.com/en-gb/eur/press/sage-ocean-announces-text-wash-as-2019-concept-grant-winner), [New Scientist II](https://www.newscientist.com/article/2223723-drill-music-with-positive-lyrics-is-more-popular-than-negative-songs/), [The Times](https://www.thetimes.co.uk/article/met-police-and-scientists-developing-way-to-decipher-voilent-threats-in-drill-videos-fzc6t9rlq), [UCL News I](https://www.ucl.ac.uk/news/headlines/2019/nov/positive-drill-music-receives-more-audience-engagement), [The Guardian](https://www.theguardian.com/world/2018/nov/02/eu-border-lie-detection-system-criticised-as-pseudoscience), [Deutsche Welle (German)](https://www.dw.com/de/eu-testet-l%C3%BCgendetektor-an-au%C3%9Fengrenzen/a-46195158), [Süddeutsche Zeitung (German)](https://www.sueddeutsche.de/digital/grenze-kuenstliche-intelligenz-software-iborderctrl-1.4196243), [Volkskrant II (Dutch)](https://www.volkskrant.nl/nieuws-achtergrond/eu-zet-omstreden-ai-leugendetector-in-bij-grenscontrole~b169bb03/), [Folia Magazine (Dutch)](https://www.folia.nl/wetenschap/126359/dialogen-met-chatbots-kunnen-helpen-in-strijd-tegen-terrorisme), [New Scientist I](https://www.newscientist.com/article/mg24032023-400-an-ai-lie-detector-will-interrogate-travellers-at-some-eu-borders/), [Volkskrant I (Dutch)](https://www.volkskrant.nl/wetenschap/een-onzindetector-voor-het-web-dat-is-nog-niet-zo-makkelijk~be927d1d/), [Psychology Today](https://www.psychologytoday.com/blog/brainstorm/201608/what-is-the-face-truth), [APA News](http://www.apa.org/news/press/releases/2016/06/judge-truthfulness.aspx) **Video and radio** [NTR Wetenschap interview](https://www.youtube.com/watch?v=C-sqX8DLs-0), [YouTube production from SAGE](https://www.youtube.com/watch?v=T9pRRn2DrMY), [BBC Newsroom](https://www.bbc.co.uk/sounds/play/w172wygv5wf8qjf), BNR Newsradio, [Amsterdam FM (Dutch)](http://www.radioswammerdam.nl/op-het-slechte-pad-i-leugentje-om-je-eigen-bestwil/) --- #### Review service ACL 2023, EMNLP 2022; EPJ Data Science; Terrorism and Political Violence; J of Policing, Intelligence and CT; ACL 2022; EMNLP 2021; Behavior Research Methods; ACL 2021; Journal of Applied Research in Memory and Cognition; EMNLP 2020; Trends & Issues in Crime and Criminal Justice; Proceedings of the National Academy of Sciences (PNAS); 54th Hawaii International Conference on System Sciences (HICSS); Scientific Reports (Nature); Psychological Science; Computers in Human Behavior; Plos One; Perspectives on Psychological Science; Affective Computing & Intelligent Interaction conference proceedings; Information; ACL 2020; AACL-IJCNLP 2020; Memory & Cognition; Applied Cognitive Psychology; Crime Science; Acta Psychologica; Personality and Individual Differences; Current Psychology; Cognitive Research: Principles and Implications; EuroCSS 2019; Computers & Security; Psychology, Crime & Law; Legal & Criminological Psychology. --- #### Software ### Software - [1] netanos - Named Entity-based Text ANonymization for Open Science, _M. Mozes_ and **B. Kleinberg**. python + node.js, https://www.npmjs.com/package/netanos, MIT license - [2] Textwash - automated, learning-based text anonymisation, _M. Mozes_ and **B. Kleinberg**. python + R, https://github.com/maximilianmozes/textwash, GPL-3 license - [3] rgpt3 - Making requests from R to the GPT-3 API, **B. Kleinberg**. R, https://github.com/ben-aaron188/rgpt3, GPL-3 license --- _This website uses the markdown [retro theme](https://github.com/markdowncss/retro) and a modified markdown-to-html pipeline._ --- If you are interested in working as a research internship (at site or remotely), a BSc/MSc project or discussing PhD options, feel free to [reach out](./sub/contact.html) to me.