PDScore: Personality driven computational psychological model for depression score estimation using social media text

Abstract

Objective

Recent studies show that the personality factors such as neuroticism, extraversion, and conscientiousness are linked to depression. This paper presents a computational model, PDScore, to explore the relationship between personality traits and depression scores obtained from an analysis of social media text by monitoring individuals’ social media posts over a specified time frame.

Method

The study aims to understand how different personality traits influence depression levels. Using the 3-D psychological theory (VAD) proposed by Mehrabian as a foundation, initial depression scores are derived. Subsequent analysis employs fuzzy logic to account for uncertainty arising from personality variability, resulting in refined depression score estimations.

Results

Validation is performed in two ways: using the eRisk2018 Dataset and Expert-Supervised Validation Using Psychometric Tools. First, the results are compared with eRisk2018 datasets and compared with the existing models. The proposed model outperformed all the existing methods with 0.84 F1-score. Then, a data is collected from 57 participants with approximately 35 to 40 posts per participant, resulting in a total of over 2150 posts. we applied our method to compute corresponding depression scores and compared them with the psychologists’ assessments. It demonstrates a correlation of 0.985 between the manual score and final depression score. The comparison yielded an accuracy of 94.7% in classification and a Root Mean Square Error (RMSE) of 0.089, indicating strong alignment and reinforcing the empirical validity of the proposed model. We consider this initial study a proof-of-concept that demonstrates the feasibility and potential of integrating personality factors and linguistic features for depression scoring.

Conclusions

The paper delves into the methodology and findings in detail, shedding light on the complex interplay between personality and mental health in online information contexts. The results reveal that personality plays an important role in estimation of depression score.

Keywords

Depression score depression detection personality traits PHQ-9 OCEAN theory

Get full access to this article

View all access options for this article.

References

Bisht

Shandilya

Gupta

, et al. Perceiving the level of depression from web text. In Artificial Intelligence, machine learning, and mental health in pandemics: a computational approach, 2022. doi:10.1016/B978-0-323-91196-2.00008-9.

Mehrabian

. Pleasure-Arousal-Dominance: a general framework for describing and measuring individual differences in temperament. Curr Psychol 1996; 14: 261–292.

Zaswita

Ihsan

. The impact of personality types on students’ writing ability. JPI (Jurnal Pendidikan Indonesia) 2020; 9: 75–84.

Tolboll

. Linguistic features in depression: A meta-analysis. J Lang Works - SprogvidenskabeligtStudentertidsskrift 2019; 4: 39–59.

Yahya

Abdul Rahim

. Linguistic markers of depression: insights from English-language tweets before and during the COVID-19 pandemic. Lang Health 2023; 1: 36–50.

Sood

Varma

Pandey

, et al. Predicting loneliness from social media text using machine learning techniques. In: Artificial Intelligence, machine learning, and mental health in pandemics: a computational approach, 2022. doi:10.1016/B978-0-323-91196-2.00011-9.

Rawat

Jain

. Emotionally wrapped social media text: approaches, opportunities, and challenges. Scalab Comput Pract Exp 2023; 24: 797–818.

Jain

Asawa

. Emotional computer: design challenges and opportunities. Int J Synthet Emot 2015; 6: 35–56.

Jain

Asawa

. Programming an expressive autonomous agent. Expert Syst Appl 2016; 43: 131–141.

10.

Saini

Bhatia

Suri

, et al. EmoXract: Domain independent emotion mining model for unstructured data. In: 2014 7th International conference on contemporary computing, IC3 2014, 2014. doi:10.1109/IC3.2014.6897154.

11.

Rawat

Jain

. A dimensional representation of depressive text. Lect Notes Data Eng Commun Technol 2021; 54: 175–187.

12.

Tausczik

Pennebaker

. The psychological meaning of words: LIWC and computerized text analysis methods. J Lang Soc Psychol 2010; 29: 24–54.

13.

Yarkoni

. Personality in 100,000 words: a large-scale analysis of personality and word use among bloggers. J Res Pers 2010; 44: 363–373.

14.

Chien

JYW

. The five-factor model of personality and depressive symptoms: one-year follow-up. Pers Individ Dif 2007; 43: 1013–1023.

15.

Akiskal

Hirschfeld

RMA

Yerevanian

. The relationship of personality to affective disorders: a critical review. Arch Gen Psychiatry 1983; 40: 101–110.

16.

Klein

Kotov

Bufferd

. Personality and depression: explanatory models and review of the evidence. Annu Rev Clin Psychol 2011; 7: 269–295.

17.

Kotov

Gamez

Schmidt

, et al. Linking ‘Big’ personality traits to anxiety, depressive, and substance use disorders: a meta-analysis. Psychol Bull 2010; 136: 768–821.

18.

Hakulinen

Elovainio

Pulkki-Råback

, et al. Personality and depressive symptoms: individual participant meta-analysis of 10 cohort studies. Depress Anxiety 2015; 32: 461–470.

19.

Zeigler-Hill

Shackelford

. The SAGE Handbook of personality and individual differences: Volume III: applications of personality and individual differences. 2018. doi:10.4135/9781526451248.

20.

Avia

Sanz

Sánchez-Bernardos

, et al. The five-factor model-II. Relations of the NEO-PI with other personality variables. Pers Individ Dif 1995; 19: 81–97.

21.

Jourdy

Petot

. Relationships between personality traits and depression in the light of the ‘Big Five’ and their different facets. Evolut Psychiatrique 2017; 82: e27–e37.

22.

Lyon

Juhasz

Brown

LJE

, et al. Big Five personality facets explaining variance in anxiety and depressive symptoms in a community sample. J Affect Disord 2020; 274: 515–521.

23.

Costa

McCrae

. Normal personality assessment in clinical practice: the NEO Personality Inventory. Psychol Assess 1992; 4: 5–13.

24.

Mehrabian

. Analysis of the big-five personality factors in terms of the PAD temperament model. Aust J Psychol 1996; 48: 86–92.

25.

Mehrabian

. Framework for a comprehensive description and measurement of emotional states. Genet Soc Gen Psychol Monogr 1995; 121: 339–361.

26.

Mehrabian

. Comparison of the PAD and PANAS as models for describing emotions and for differentiating anxiety from depression. J Psychopathol Behav Assess 1997; 19: 331–357.

27.

Ching

. Fuzzy logic: the genius of lofti zadeh [my view]. IEEE Ind Electron Mag 2017; 11: 6–37.

28.

Digman

. Personality structure: emergence of the five-factor model. Annu Rev Psychol 1990; 41: 417–440.

29.

Levis

Riehm

, et al. The accuracy of the patient health questionnaire-9 algorithm for screening to detect major depression: an individual participant data meta-analysis. Psychother Psychosom 2020; 89: 25–37.

30.

Asokan

Geethapriya

Dhanabalan

, et al. Assessment of personality traits among pediatric dentists in India: a cross-sectional study. Int J Clin Pediatr Dent 2023; 16: 489–493.

31.

Ramiandrisoa

Mothe

Benamara

, et al. IRIT at e-Risk 2018. In: CEUR workshop proceedings, 2018.

32.

Rawat

Jain

. Depression score estimation using dimensional theory of emotion: a web text based computation model. Under Review.

33.

Rawat

Jain

. DPre: effective preprocessing techniques for social media depressive text. Intell Decis Technol 2022; 16: 475–485.

34.

Mohammad

Turney

. Emotions evoked by common words and phrases: using mechanical turk to create an emotion lexicon. In: CAAGET ‘10 Proceedings of the NAACL HLT 2010 workshop on computational approaches to analysis and generation of emotion in text, 2010.

35.

Udovičić

. Big five personality traits and social support as predictors of postpartum depression. J Eur Psychol Students 2014; 5: 66–73.

36.

Lester

. Depression, suicidal ideation and the big five personality traits. Aust J Psychiatry Behav Sci 2021; 7: 1–4.

37.

Koorevaar

AML

Comijs

Dhondt

ADF

, et al. Big Five personality and depression diagnosis, severity and age of onset in older adults. J Affect Disord 2013; 151: 178–185.

38.

Steunenberg

Braam

Beekman

ATF

, et al.

Evidence for an association of the big five personality factors with recurrence of depressive symptoms in later life.

Int J Geriatr Psychiatry 2009; 24: 1470–1477.

39.

Koorevaar

AML

Hegeman

Lamers

, et al. Big Five personality characteristics are associated with depression subtypes and symptom dimensions of depression in older adults. Int J Geriatr Psychiatry 2017; 32: e132–e140.

40.

Leow

Lee

Lynch

. Big five personality and depressive symptoms: a Self-Determination Theory perspective on students’ positive relationships with others. Ideas Res You Can Use: VISTAS 2016; 2016: 2–11.

41.

Naito

Kato

Koshikawa

, et al. Personality as a basis for antidepressant selection for patients with depression: a two-point outcome study at 4 and 8 weeks. J Affect Disord 2022; 314: 27–33.

42.

Papastylianou

. Relating on the internet, personality traits and depression: research and implications. Eur J Counsell Psychol 2013; 2: 65–78.

43.

Nouri

Feizi

Afshar

, et al. How five-factor personality traits affect psychological distress and depression? Results from a large population-based study. Psychol Stud (Mysore) 2019; 64: 59–69.

44.

Funez

Ucelay

MJG

Villegas

, et al. UNSL’s participation at eRisk 2018 Lab. In: Working Notes of CLEF 2018 - conference and labs of the evaluation forum, Avignon, France, September 10–14, 2018.

45.

Trotzek

Koitka

Friedrich

. Word Embed- dings and Linguistic Metadata at the CLEF 2018 Tasks for Early Detection of Depression and Anorexia. In: Working Notes of CLEF 2018 - conference and labs of the evaluation forum, Avignon, France, September 10–14, 2018.

46.

Ramiandrisoa

Mothe

Benamara

, et al. IRIT at e-Risk 2018. In Experimental IR Meets Multilinguality, Multimodal- ity, and Interaction. Proceedings of the Ninth International Conference of the CLEF Association (CLEF 2018), Avignon, France, 2018.

47.

Ramiandrisoa

Mothe

. Early detection of depression and anorexia from social media: a machine learning approach. Circle 2020, Iván Cantador; Max Chevalier; Massimo Melucci; Josiane Mothe, Jul 2020, Samatan, France. hal-02877723.