Online reading habits can reveal personality traits: Towards detecting psychological microtargeting

Almog Simchon; Adam J Sutton; Matthew  Edwards; Stephan Lewandowsky

doi:10.1093/pnasnexus/pgad191

Online reading habits can reveal personality traits: Towards detecting psychological microtargeting

Almog Simchon^*, Adam J Sutton, Matthew Edwards, Stephan Lewandowsky

^*Corresponding author for this work

School of Psychological Science

Research output: Contribution to journal › Article (Academic Journal) › peer-review

11 Citations (Scopus)

Abstract

Building on big data from Reddit, we generated two computational text models: (1) Predicting the personality of users from the text they have written and (2) predicting the personality of users based on the text they have consumed. The second model is novel and without precedent in the literature. We recruited active Reddit users (N = 1,105) of fiction-writing communities. The participants completed a Big Five personality questionnaire, and consented for their Reddit activity to be scraped and used to create a machine-learning model. We trained an NLP model (BERT), predicting personality from produced text (average performance: r = 0.33). We then applied this model to a new set of Reddit users (N = 10,050), predicted their personality based on their produced text, and trained a second BERT model to predict their predicted-personality scores based on consumed text (average performance: r = 0.13). By doing so, we provide the first glimpse into the linguistic markers of personality-congruent consumed content.

Original language	English
Article number	pgad191
Journal	PNAS Nexus
Volume	2
Issue number	6
Early online date	7 Jun 2023
DOIs	https://doi.org/10.1093/pnasnexus/pgad191
Publication status	Published - 17 Jun 2023

Bibliographical note

Funding Information:
This research was supported by a large grant from the Volkswagen Foundation (“Reclaiming individual autonomy and democratic discourse online”). S.L. was also supported by funding from the Humboldt Foundation in Germany and by an Advanced Grant (PRODEMINFO). A.Su. was also supported by funding by REPHRAIN: National Research Centre on Privacy, Harm Reduction and Adversarial Influence online (EPSRC Grant: EP/V011189/1).

Funding Information:
The authors thank Danae Ioakeimidou and the anonymous reviewers for their valuable suggestions. This manuscript was posted on a preprint: https://osf.io/wdrus. This research was supported by a large grant from the Volkswagen Foundation (“Reclaiming individual autonomy and democratic discourse online”). S.L. was also supported by funding from the Humboldt Foundation in Germany and by an Advanced Grant (PRODEMINFO). A.Su. was also supported by funding by REPHRAIN: National Research Centre on Privacy, Harm Reduction and Adversarial Influence online (EPSRC Grant: EP/V011189/1).

Publisher Copyright:
© The Author(s) 2023. Published by Oxford University Press on behalf of National Academy of Sciences. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

Research Groups and Themes

Cyber Security
Cognitive Science
TeDCog

Access to Document

10.1093/pnasnexus/pgad191Licence: CC BY

Handle.net

Persistent link

Protecting the Democratic Information Space in Europe
Lewandowsky, S. (Principal Investigator), Westaway, R. M. (Administrator) & Carrella, F. (Researcher)
1/10/21 → 30/09/26
Project: Research, Parent

Cite this

@article{9d7ac7099d314f7eb619c943be6cbbb0,

title = "Online reading habits can reveal personality traits: Towards detecting psychological microtargeting",

abstract = "Building on big data from Reddit, we generated two computational text models: (1) Predicting the personality of users from the text they have written and (2) predicting the personality of users based on the text they have consumed. The second model is novel and without precedent in the literature. We recruited active Reddit users (N = 1,105) of fiction-writing communities. The participants completed a Big Five personality questionnaire, and consented for their Reddit activity to be scraped and used to create a machine-learning model. We trained an NLP model (BERT), predicting personality from produced text (average performance: r = 0.33). We then applied this model to a new set of Reddit users (N = 10,050), predicted their personality based on their produced text, and trained a second BERT model to predict their predicted-personality scores based on consumed text (average performance: r = 0.13). By doing so, we provide the first glimpse into the linguistic markers of personality-congruent consumed content.",

author = "Almog Simchon and Sutton, \{Adam J\} and Matthew Edwards and Stephan Lewandowsky",

note = "Funding Information: This research was supported by a large grant from the Volkswagen Foundation (“Reclaiming individual autonomy and democratic discourse online”). S.L. was also supported by funding from the Humboldt Foundation in Germany and by an Advanced Grant (PRODEMINFO). A.Su. was also supported by funding by REPHRAIN: National Research Centre on Privacy, Harm Reduction and Adversarial Influence online (EPSRC Grant: EP/V011189/1). Funding Information: The authors thank Danae Ioakeimidou and the anonymous reviewers for their valuable suggestions. This manuscript was posted on a preprint: https://osf.io/wdrus. This research was supported by a large grant from the Volkswagen Foundation (“Reclaiming individual autonomy and democratic discourse online”). S.L. was also supported by funding from the Humboldt Foundation in Germany and by an Advanced Grant (PRODEMINFO). A.Su. was also supported by funding by REPHRAIN: National Research Centre on Privacy, Harm Reduction and Adversarial Influence online (EPSRC Grant: EP/V011189/1). Publisher Copyright: {\textcopyright} The Author(s) 2023. Published by Oxford University Press on behalf of National Academy of Sciences. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.",

year = "2023",

month = jun,

day = "17",

doi = "10.1093/pnasnexus/pgad191",

language = "English",

volume = "2",

journal = "PNAS Nexus",

issn = "2752-6542",

publisher = "National Academy of Sciences",

number = "6",

}

TY - JOUR

T1 - Online reading habits can reveal personality traits

T2 - Towards detecting psychological microtargeting

AU - Simchon, Almog

AU - Sutton, Adam J

AU - Edwards, Matthew

AU - Lewandowsky, Stephan

N1 - Funding Information: This research was supported by a large grant from the Volkswagen Foundation (“Reclaiming individual autonomy and democratic discourse online”). S.L. was also supported by funding from the Humboldt Foundation in Germany and by an Advanced Grant (PRODEMINFO). A.Su. was also supported by funding by REPHRAIN: National Research Centre on Privacy, Harm Reduction and Adversarial Influence online (EPSRC Grant: EP/V011189/1). Funding Information: The authors thank Danae Ioakeimidou and the anonymous reviewers for their valuable suggestions. This manuscript was posted on a preprint: https://osf.io/wdrus. This research was supported by a large grant from the Volkswagen Foundation (“Reclaiming individual autonomy and democratic discourse online”). S.L. was also supported by funding from the Humboldt Foundation in Germany and by an Advanced Grant (PRODEMINFO). A.Su. was also supported by funding by REPHRAIN: National Research Centre on Privacy, Harm Reduction and Adversarial Influence online (EPSRC Grant: EP/V011189/1). Publisher Copyright: © The Author(s) 2023. Published by Oxford University Press on behalf of National Academy of Sciences. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

PY - 2023/6/17

Y1 - 2023/6/17

N2 - Building on big data from Reddit, we generated two computational text models: (1) Predicting the personality of users from the text they have written and (2) predicting the personality of users based on the text they have consumed. The second model is novel and without precedent in the literature. We recruited active Reddit users (N = 1,105) of fiction-writing communities. The participants completed a Big Five personality questionnaire, and consented for their Reddit activity to be scraped and used to create a machine-learning model. We trained an NLP model (BERT), predicting personality from produced text (average performance: r = 0.33). We then applied this model to a new set of Reddit users (N = 10,050), predicted their personality based on their produced text, and trained a second BERT model to predict their predicted-personality scores based on consumed text (average performance: r = 0.13). By doing so, we provide the first glimpse into the linguistic markers of personality-congruent consumed content.

AB - Building on big data from Reddit, we generated two computational text models: (1) Predicting the personality of users from the text they have written and (2) predicting the personality of users based on the text they have consumed. The second model is novel and without precedent in the literature. We recruited active Reddit users (N = 1,105) of fiction-writing communities. The participants completed a Big Five personality questionnaire, and consented for their Reddit activity to be scraped and used to create a machine-learning model. We trained an NLP model (BERT), predicting personality from produced text (average performance: r = 0.33). We then applied this model to a new set of Reddit users (N = 10,050), predicted their personality based on their produced text, and trained a second BERT model to predict their predicted-personality scores based on consumed text (average performance: r = 0.13). By doing so, we provide the first glimpse into the linguistic markers of personality-congruent consumed content.

U2 - 10.1093/pnasnexus/pgad191

DO - 10.1093/pnasnexus/pgad191

M3 - Article (Academic Journal)

C2 - 37333766

SN - 2752-6542

VL - 2

JO - PNAS Nexus

JF - PNAS Nexus

IS - 6

M1 - pgad191

ER -

Online reading habits can reveal personality traits: Towards detecting psychological microtargeting

Abstract

Bibliographical note

Research Groups and Themes

Access to Document

Handle.net

Fingerprint

Projects

Protecting the Democratic Information Space in Europe

Cite this