HEAL-Summ: a lightweight and ethical framework for accessible summarization of health information
Bibliographic record
Abstract
Introduction: The growing volume and complexity of health-related news presents significant barriers to public understanding. While large language models (LLMs) offer a promising means of summarizing such content, many approaches are computationally expensive and can lack sufficient evaluation of ethical as well as representational quality. Methods: To address these limitations, this research proposes a lightweight framework called Health Ethics & Accessibility with Lightweight Summarization (HEAL-Summ) for summarizing Canadian health news articles using LLMs. The framework incorporates three models (Phi 3, Qwen 2.5, and Llama 3.2) and integrates a multi-dimensional evaluation strategy to assess semantic consistency, readability, lexical diversity, emotional alignment, and toxicity. Results: Comparative analyses shows consistent semantic agreement across models, with Phi yielding more accessible summaries and Qwen producing greater emotional as well as lexical diversity. Statistical significance testing supports key differences in readability and emotional tone. Discussion: This work goes beyond single-model summarization by providing a structured and ethical framework for longitudinal news analysis, emphasizing low-resource deployment and built-in automated evaluations. The findings highlight the potential for lightweight LLMs to facilitate transparent and emotionally sensitive communication in public health, while maintaining a balance between linguistic expressiveness and ethical reliability. The proposed framework offers a scalable path forward for improving access to complex health information in resource-constrained or high-stakes environments.
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 | 0.011 | 0.057 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.002 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.005 | 0.004 |
| Open science | 0.002 | 0.004 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.010 | 0.003 |
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".