Automated Credibility Assessment of Web-Based Health Information Considering Health on the Net Foundation Code of Conduct (HONcode): Model Development and Validation Study
Bibliographic record
Abstract
BACKGROUND: An increasing number of users are turning to web-based sources as an important source of health care guidance information. Thus, trustworthy sources of information should be automatically identifiable using objective criteria. OBJECTIVE: The purpose of this study was to automate the assessment of the Health on the Net Foundation Code of Conduct (HONcode) criteria, enhancing our ability to pinpoint trustworthy health information sources. METHODS: A data set of 538 web pages displaying health content was collected from 43 health-related websites. HONcode criteria have been considered as web page and website levels. For the website-level criteria (confidentiality, transparency, financial disclosure, and advertising policy), a bag of keywords has been identified to assess the criteria using a rule-based model. For the web page-level criteria (authority, complementarity, justifiability, and attribution) several machine learning (ML) approaches were used. In total, 200 web pages were manually annotated until achieving a balanced representation in terms of frequency. In total, 3 ML models-random forest, support vector machines (SVM), and Bidirectional Encoder Representations from Transformers (BERT)-were trained on the initial annotated data. A second step of training was implemented for the complementarity criterion using the BERT model for multiclass classification of the complementarity sentences obtained by annotation and data augmentation (positive, negative, and noncommittal sentences). Finally, the remaining web pages were classified using the selected model and 100 sentences were randomly selected for manual review. RESULTS: For web page-level criteria, the random forest model showed a good performance for the attribution criterion while displaying subpar performance in the others. BERT and SVM had a stable performance across all the criteria. BERT had a better area under the curve (AUC) of 0.96, 0.98, and 1.00 for neutral sentences, justifiability, and attribution, respectively. SVM had the overall better performance for the classification of complementarity with the AUC equal to 0.98. Finally, SVM and BERT had an equal AUC of 0.98 for the authority criterion. For the website level criteria, the rule-based model was able to retrieve web pages with an accuracy of 0.97 for confidentiality, 0.82 for transparency, and 0.51 for both financial disclosure and advertising policy. The final evaluation of the sentences determined 0.88 of precision and the agreement level of reviewers was computed at 0.82. CONCLUSIONS: Our results showed the potential power of automating the HONcode criteria assessment using ML approaches. This approach could be used with different types of pretrained models to accelerate the text annotation, and classification and to improve the performance in low-resource cases. Further work needs to be conducted to determine how to assign different weights to the criteria, as well as to identify additional characteristics that should be considered for consolidating these criteria into a comprehensive reliability score.
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 | 0.013 | 0.033 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 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".