MétaCan
Menu
Back to cohort
Record W7135184301 · doi:10.2196/74397

Emotional Expression and Mental Health Support in BTS Fandom Communities: A Natural Language Processing Study on YouTube Comments (Preprint)

2025· article· en· W7135184301 on OpenAlexvenueno aff
Nari Yoo, Aaron H. Rodwin, Michael Park, Sangpil Youm, Sou Hyun Jang

Bibliographic record

VenueJMIR Infodemiology · 2025
Typearticle
Languageen
FieldPsychology
TopicMental Health via Writing
Canadian institutionsnot available
Fundersnot available
KeywordsMental healthExpression (computer science)FandomEmotional expressionSocial mediaNatural (archaeology)

Abstract

fetched live from OpenAlex

BACKGROUND: The global rise of K-pop has shaped youth culture and online communities, particularly through BTS, a South Korean boy band with an international fanbase known as ARMY (Adorable Representative MC for Youth). Music fandoms are increasingly engaging with digital platforms such as YouTube not only for entertainment but also as spaces for emotional expression and mutual support. Despite growing interest in the mental health potential of music-based coping strategies, limited research has examined how fandom cultures differentially express emotional needs and supportive interactions online. OBJECTIVE: This study investigates specific mental health language patterns and coping mechanisms expressed by BTS fans in online spaces, examining how different linguistic features (including self-referential language and emotional expression patterns) may reflect psychological states and mental health needs. We used YouTube comments from fan-curated "sad" or "depression" playlists of BTS. We further included YouTube comments from equivalent Taylor Swift playlists as a reference group. METHODS: Using natural language processing and Linguistic Inquiry and Word Count 2022 software, we analyzed a total of 13,224 YouTube comments: 11,772 comments on BTS "sad playlist" videos and 1452 comments on Taylor Swift equivalents. Statistical comparisons were conducted to evaluate differences in comment length, word count, pronoun use, and emotional valence. Representative comments were examined to contextualize the emotion classification results. RESULTS: BTS original comments were significantly longer (mean 253.38, SD 703.65 characters) and had higher word counts (mean 38.93, SD 88.54 words) than Taylor Swift original comments (length: mean 89.84, SD 330.96 characters; word count: mean 16.08, SD 64.78 words; P<.001). BTS fans used more first-person singular pronouns (mean 10.24%, SD 9.57% vs mean 7.43%, SD 9.41%) and expressed greater sadness (1691/5341, 31.7% vs 75/452, 16.6%). In contrast, Taylor Swift fans exhibited higher admiration (86/452, 19% vs 429/5341, 8%). Among reply comments, BTS fans demonstrated more caring (242/1729, 14% vs 7/128, 5.5%), gratitude (294/1729, 17% vs 15/128, 11.7%), and optimism (162/1729, 9.4% vs 6/128, 4.7%). Linguistic analysis also revealed a broader international user base for BTS, including higher proportions of Spanish (719/11,772, 6.11%) and Portuguese (222/11,772, 1.89%) comments. Examination of comment content showed that fans used these spaces to disclose personal struggles, express gratitude for the community, and offer peer support, with many describing the fandom as a safe space for emotional expression they could not access elsewhere. CONCLUSIONS: The findings show that comments on BTS fan playlists included more emotional disclosure and more supportive replies than those on the Taylor Swift comparison, consistent with fan communities operating as informal sites of peer support online. These results suggest implications for culturally responsive, community-based, and digitally mediated mental health interventions among youth and global populations.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.009
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.020
Threshold uncertainty score0.040

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0010.009
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0020.002
Science and technology studies0.0010.001
Scholarly communication0.0010.001
Open science0.0000.001
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0020.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.045
GPT teacher head0.455
Teacher spread0.411 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designSimulation or modeling
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations1
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueJMIR InfodemiologySame topicMental Health via WritingFrench-language works237,207