Developing a standardized framework for evaluating health apps using natural language processing
Bibliographic record
Abstract
Despite regulatory efforts, many smartphone health applications remain unregulated, raising concerns about privacy, security, and evidence-based effectiveness. The lack of standardized regulation has led to the proliferation of over 130 frameworks, introducing new criteria and methodologies for app evaluation. The sheer number of frameworks, coupled with their varying approaches to app evaluation, create challenges for comparison. Our study aims to synthesize existing knowledge and propose a standardized app evaluation framework. We conducted a synthesis of reviews on health app evaluation frameworks. Using natural language processing (NLP), we analyzed evaluation domains and grouped them into clusters based on semantic similarities. Standardized definitions for these clusters were developed. We identified eight review articles that met the inclusion criteria, each proposing between six and 17 app evaluation domains. Using NLP, we identified five clusters of app evaluation: Effectiveness & Development, Technology & Functionality, Validity & Legal, Safety & Privacy, and Implementation & Ethics, each of which was assigned a standardized definition. The clusters align with but expand on the American Psychiatric Association's evaluation domains, incorporating critical aspects such as inclusivity, safety, engagement, and ethical principles. Temporal analysis revealed an increasing focus on Effectiveness & Development, while Safety & Privacy showed a stagnation in attention over time.
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 | 0.325 | 0.405 |
| Meta-epidemiology (narrow) | 0.005 | 0.002 |
| Meta-epidemiology (broad) | 0.005 | 0.013 |
| Bibliometrics | 0.050 | 0.020 |
| Science and technology studies | 0.004 | 0.007 |
| Scholarly communication | 0.017 | 0.014 |
| Open science | 0.006 | 0.012 |
| Research integrity | 0.005 | 0.005 |
| Insufficient payload (model declined to judge) | 0.005 | 0.002 |
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; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
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".