Digital Disease Surveillance for Emerging Infectious Diseases: An Early Warning System Using the Internet and Social Media Data for COVID-19 Forecasting in Canada
Bibliographic record
Abstract
BACKGROUND: Emerging Infectious Diseases (EID) are a significant threat to population health globally. We aimed to examine the relationship between internet search engine queries and social media data on COVID-19 and determine if they can predict COVID-19 cases in Canada. METHODS: We analyzed Google Trends (GT) and Twitter data from 1/1/2020 to 3/31/2020 in Canada and used various signal-processing techniques to remove noise from the data. Data on COVID-19 cases was obtained from the COVID-19 Canada Open Data Working Group. We conducted time-lagged cross-correlation analyses and developed the long short-term memory model for forecasting daily COVID-19 cases. RESULTS: Among symptom keywords, "cough," "runny nose," and "anosmia" were strong signals with high cross-correlation coefficients >0.8 ( rCough = 0.825, t - 9; rRunnyNose = 0.816, t - 11; rAnosmia = 0.812, t - 3 ), showing that searching for "cough," "runny nose," and "anosmia" on GT correlated with the incidence of COVID-19 and peaked 9, 11, and 3 days earlier than the incidence peak, respectively. For symptoms- and COVID-related Tweet counts, the cross-correlations of Tweet signals and daily cases were rTweetSymptoms = 0.868, t - 11 and tTweetCOVID = 0.840, t - 10, respectively. The LSTM forecasting model achieved the best performance (MSE = 124.78, R2 = 0.88, adjusted R2 = 0.87) using GT signals with cross-correlation coefficients >0.75. Combining GT and Tweet signals did not improve the model performance. CONCLUSION: Internet search engine queries and social media data can be used as early warning signals for creating a real-time surveillance system for COVID-19 forecasting, but challenges remain in modelling.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.005 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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.
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 teacher head, 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".