An Analysis of the Prevalence and Trends in Drug-Related Lyrics on Twitter (X): Quantitative Approach
Bibliographic record
Abstract
BACKGROUND: The pervasiveness of drug culture has become evident in popular music and social media. Previous research has examined drug abuse content in both social media and popular music; however, to our knowledge, the intersection of drug abuse content in these 2 domains has not been explored. To address the ongoing drug epidemic, we analyzed drug-related content on Twitter (subsequently rebranded X), with a specific focus on lyrics. Our study provides a novel finding on the prevalence of drug abuse by defining a new subcategory of X content: "tweets that reference established drug lyrics." OBJECTIVE: We aim to investigate drug trends in popular music on X, identify and classify popular drugs, and analyze related artists' gender, genre, and popularity. Based on the collected data, our goal is to create a prediction model for future drug trends and gain a deeper understanding of the characteristics of users who cite drug lyrics on X. METHODS: X data were collected from 2015 to 2017 through the X streaming application programming interface (API). Drug lyrics were obtained from the Genius lyrics database using the Genius API based on drug keywords. The Smith-Waterman text-matching algorithm is used to detect the drug lyrics in posts. We identified famous drugs in lyrics that were posted. Consequently, the analysis was extended to related artists, songs, genres, and popularity on X. The frequency of drug-related lyrics on X was aggregated into a time-series, which was then used to create prediction models using linear regression, Facebook Prophet, and NIXTLA TimeGPT-1. In addition, we analyzed the number of followers of users posting drug-related lyrics to explore user characteristics. RESULTS: We analyzed over 1.97 billion publicly available posts from 2015 to 2017, identifying more than 157 million that matched drug-related keywords. Of these, 150,746 posts referenced drug-related lyrics. Cannabinoids, opioids, stimulants, and hallucinogens were the most cited drugs in lyrics on X. Rap and hip-hop dominated, with 91.98% of drug-related lyrics from these genres and 84.21% performed by male artists. Predictions from all 3 models, linear regression, Facebook Prophet, and NIXTLA TimeGPT-1, indicate a slight decline in the prevalence of drug-related lyrics on X over time. CONCLUSIONS: Our study revealed 2 significant findings. First, we identified a previously unexamined subset of drug-related content on X: drug lyrics, which could play a critical role in models predicting the surge in drug-related incidents. Second, we demonstrated the use of cutting-edge time-series forecasting tools, including Facebook Prophet and NIXTLA TimeGPT-1, in accurately predicting these trends. These insights contribute to our understanding of how social media shapes public behavior and sentiment toward drug use.
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 imitationNot 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.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.008 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.006 | 0.006 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".