cs.CLFeb 20, 2026

PsihoRo: Depression and Anxiety Romanian Text Corpus

Authors: Alexandra Ciobotaru, Ana-Maria Bucur, Liviu P. Dinu

Organizations: University of Bucharest, 490 Panduri Street, Bucharest, Romania · Druid AI, 1-5 Costache Negri Street, Bucharest, Romania · Università della Svizzera italiana, Via Buffi 13, Lugano, Switzerland · Human Language Technologies Research Center

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Psychological corpora in NLP are collections of texts used to analyze human psychology, emotions, and mental health. These texts allow researchers to study psychological constructs, identify patterns related to mental health problems and analyze emotional language. However, collecting accurate mental health data from social media can be challenging due to the assumptions made by data collectors. A more effective approach involves gathering data through open-ended questions and then assessing participants' mental health status using self-report screening surveys. This method was successfully employed for English, a language with a lot of psychological NLP resources. However, the same cannot be stated for Romanian, which currently has no open-source mental health corpus. To address this gap, we have collected the first open-source corpus focused on depression and anxiety in Romanian, by utilizing a form with 6 open-ended questions along with the standardized PHQ-9 and GAD-7 screening questionnaires. Although the PsihoRo corpus contains texts from only 205 respondents, it represents an important first step toward understanding and analyzing mental health issues within the Romanian population. We employ statistical analysis, text analysis using Romanian LIWC, emotion detection, and topic modeling to identify the most important features of this newly introduced resource for the NLP community. The data is publicly available at https://huggingface.co/datasets/Alegzandra/PsihoRo.

Figures & tables

 PHQ-9GAD-7
Count205205
Mean6.005.78
SD4.834.95
Min00
25%32
50% (Median)55
Table 1: Descriptive statistics for PsihoRo
Figure 1: Distributions for GAD-7 and PHQ-9
Figure 2: Correlation between Anxiety and Depression Scores.
Figure 3: Differences in selected LIWC linguistic categories between participants with low vs. high PHQ-9 (left) and GAD-7 (right) scores. * denotes statistically significant differences ( p<0.05p<0.05 ) measured using the Mann-Whitney U test. Positive values indicate greater relative use among participants with higher symptom scores.
Figure 4: Feature importance for LIWC categories analyzed using SHAP.
Figure 5: Emotion predictions across: all PsihoRo texts (left), negative-topic (pessimistic) questions (middle), and positive-topic (optimistic) questions (right).
Figure 6: Emotion predictions across the three studied population groups.
More in this topic

Clinical NLP

Clinical natural language processing (NLP) develops methods to extract, interpret, and generate information from healthcare text and conversations.

See all 142 papers

Explore similar work

CardsList

  1. Leveraging LLMs for Translating and Classifying Mental Health Data

    Oct 16, 2024Konstantinos Skianis, A. Seza Doğruöz, John PavlopoulosDepression Severity EstimationClinical NLP

    Department of Computer Science and Engineering, University of Ioannina, Greece · LT3, IDLab, Universiteit Gent, Belgium · Department of Informatics, Athens University of Economics and Business, Greece +1

  2. Mental Health Disorder Detection Beyond Social Media: A Systematic Review of Available Datasets

    Jul 3, 2026Sadiya Sayara Chowdhury Puspo, Ana-Maria Bucur, Stevie Chancellor +2Mental Health

    George Mason University, USA · Università della Svizzera italiana, Switzerland · University of Minnesota, USA

  3. Natural Language Processing Psychometrics

    Aug 7, 2026Edoardo Sebastiano De Duro, Emma Franchino, Massimo StellaClinical Outcome PredictionPsychometric Validation

    CogNoscoLab, Department of Psychology and Cognitive Science, University of Trento, Italy