Contextual recommendation modeling in eCoaching with machine learning, X-AI, and semantic ontology - Report - MDSpire

Contextual recommendation modeling in eCoaching with machine learning, X-AI, and semantic ontology

  • By

  • Ayan Chatterjee

  • Nurilla Avazov

  • July 15, 2026

Share

Clinical Report: Modeling Contextual Recommendations in eCoaching Using Machine Learning

Background

Sedentary lifestyles are prevalent globally, contributing to various health risks such as obesity and cardiovascular diseases. Weather conditions significantly impact physical activity levels, often discouraging outdoor exercise. An eCoaching system that offers personalized recommendations based on weather data can help mitigate these barriers.

Data Highlights

MetricValue
Accuracy99.1%

Key Findings

  • The eCoaching system uses real-time weather data to generate tailored exercise recommendations.
  • Weather data was collected over 18 months from thirteen cities in southern Norway.
  • The decision tree classifier demonstrated a high accuracy of 99.1% in classifying suitable activity types.
  • Local model-independent interpretable explanations (LIME) were utilized to enhance the interpretability of predictions.
  • The system's ontology framework provides a reliable semantic representation for activity recommendations.

Clinical Implications

Healthcare providers can utilize this eCoaching system to provide personalized activity guidance to patients during adverse weather conditions.

Conclusion

The development of a weather-aware eCoaching system aims to promote physical activity through personalized recommendations based on real-time weather data.

Related Resources & Content

  1. CDC, Physical Activity Basics, 2026 -- What You Can Do to Meet Physical Activity Recommendations
  2. npj Digital Medicine, 2025 -- Systematic review and meta analysis of standalone digital behavior change interventions on physical activity
  3. Frontiers in Digital Health — Large language models for promoting physical activity: a review of experiential and behavioral outcomes, social roles, and human-likeness in persuasive LLMs
  4. npj Digital Medicine — Addressing Mentorship Disparities: The Potential of Large Language Models to Transform Equity in the Medical Workforce
  5. npj Digital Medicine — Enhanced Transferability of Predictions from Electronic Health Records Across Different Countries and Coding Frameworks Using Large Language Models
  6. JMIR Medical Informatics — Clinical Context Variables Collectively Rival Model Choice in Embedding-Based Retrieval: Multi-Corpus Benchmark Study
  7. What You Can Do to Meet Physical Activity Recommendations | Physical Activity Basics | CDC
  8. Systematic review and meta analysis of standalone digital behavior change interventions on physical activity | npj Digital Medicine
  9. Ethics and governance of artificial intelligence for health: Guidance on large multi-modal models

Original Source(s)

Contextual recommendation modeling in eCoaching with machine learning, X-AI, and semantic ontology

  • by Ayan Chatterjee, Nurilla Avazov

  • July 15, 2026

Related Content

Autism Spectrum Disorder Linked to Distinct Urine Metabolite Pattern

Researchers proposed a new autism spectrum disorder phenotype characterized by elevated urinary microbially derived metabolites.

Does High PCT Warrant Longer Therapy?

In a planned BALANCE analysis, elevated day 7 procalcitonin was associated with higher mortality, but the small subgroup analysis found no evidence that extending antibiotics from 7 to 14 days improved outcomes.

Insurance Type and Menopausal Hormone Therapy Use Among US Women