Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Classification of Illness01:17

Classification of Illness

9.0K
The meaning of illness is individualized to each person who experiences an alteration in health. In contrast, disease is a medical term indicating a pathological change in the structure and function of the body or mind. It is a condition that has specific symptoms and boundaries.
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...
9.0K
Sex-linked Disorders01:43

Sex-linked Disorders

109.6K
Like autosomes, sex chromosomes contain a variety of genes necessary for normal body function. When a mutation in one of these genes results in biological deficits, the disorder is considered sex-linked.
109.6K
How Data are Classified: Categorical Data01:11

How Data are Classified: Categorical Data

45.9K
A variable, usually notated by capital letters such as X and Y, is a characteristic or measurement that can be determined for each member of a population. Data are the actual values of variables. They may be numbers, or they may be words. Datum is a single value.
Data are classified based on whether they are measurable or not. Categorical data cannot be measured; instead, it can be divided into categories. For example, if Y denotes a person's party affiliation, some examples of Y include...
45.9K
Classification of Systems-II01:31

Classification of Systems-II

530
Continuous-time systems have continuous input and output signals, with time measured continuously. These systems are generally defined by differential or algebraic equations. For instance, in an RC circuit, the relationship between input and output voltage is expressed through a differential equation derived from Ohm's law and the capacitor relation,
530
Levels of Health Promotion and Illness Prevention01:26

Levels of Health Promotion and Illness Prevention

15.0K
Health promotion allows a person to control the determinants of health, resulting in an improved health status. It enhances the quality of life and reduces premature deaths. Health promotion and illness prevention programs help people make beneficial choices to reduce the risk of disease and disabilities. There are three health promotion and illness prevention levels: primary, secondary, and tertiary prevention.
In primary prevention, actions taken before disease onset prevent the disease from...
15.0K
Classification of Signals01:30

Classification of Signals

1.5K
In signal processing, signals are classified based on various characteristics: continuous-time versus discrete-time, periodic versus aperiodic, analog versus digital, and causal versus noncausal. Each category highlights distinct properties crucial for understanding and manipulating signals.
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
1.5K

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Promise to Practice: Reimagining Artificial Intelligence for Equitable Global Health Impact.

Annals of global health·2026
Same author

AI-assisted teams outperform AI-led teams but not human-only teams in assessing research reproducibility in quantitative social science.

Proceedings of the National Academy of Sciences of the United States of America·2026
Same author

Understanding User Intent in Code-Mixed Sexual and Reproductive Health Queries in Urban India: Hierarchical Classification Approach Using Large Language Models.

Journal of medical Internet research·2026
Same author

Inference Gap in Domain Expertise and Machine Intelligence in Named Entity Recognition: Creation of and Insights from a Substance Use-related Dataset.

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing·2026
Same author

"Now that they come to our doorsteps to teach us these things…" - Postpartum contraception outcomes from a pre-post effectiveness-implementation study of an integrated community health worker intervention in rural Nepal.

Reproductive health·2025
Same author

Using a non-surgical transcutaneous intraglandular injection technique to deliver cell and cell-free therapies to murine submandibular salivary glands.

PloS one·2025

Related Experiment Video

Updated: Feb 22, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

1.2K

Understanding User Intent in Code-Mixed Sexual and Reproductive Health Queries in Urban India: A Hierarchical

Sumon Kanti Dey1, Manvi S1, Aradhana Thapa2

  • 1Department of Biomedical Informatics, School of Medicine, Emory University, 101 Woodruff Circle, Atlanta, US.

Journal of Medical Internet Research
|February 20, 2026
PubMed
Summary

Large language models (LLMs) show promise for sexual and reproductive health (SRH) queries in Hinglish. Proprietary models lead, but Indic LLMs like Sarvam-M demonstrate strong potential for culturally relevant AI in underserved communities.

More Related Videos

Project-Based Learning Guidelines for Health Sciences Students: An Analysis with Data Mining and Qualitative Techniques
13:44

Project-Based Learning Guidelines for Health Sciences Students: An Analysis with Data Mining and Qualitative Techniques

Published on: December 9, 2022

4.6K

Related Experiment Videos

Last Updated: Feb 22, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

1.2K
Project-Based Learning Guidelines for Health Sciences Students: An Analysis with Data Mining and Qualitative Techniques
13:44

Project-Based Learning Guidelines for Health Sciences Students: An Analysis with Data Mining and Qualitative Techniques

Published on: December 9, 2022

4.6K

Area of Science:

  • Artificial Intelligence
  • Natural Language Processing
  • Global Health

Background:

  • Sexual and reproductive health (SRH) information access is limited globally, especially in the Global South, due to stigma and linguistic barriers.
  • Code-mixed languages like Hinglish and colloquial terms are common in SRH queries in India, posing challenges for standard AI systems.
  • Existing large language models (LLMs) often perform poorly on non-English, code-mixed, and culturally specific content.

Purpose of the Study:

  • To evaluate the effectiveness of proprietary, multilingual open-weight, and Indic LLMs in understanding user intent for Hinglish SRH queries in zero-shot settings.
  • To assess LLM performance in a two-level hierarchical classification of SRH topics and subtopics, crucial for accurate guidance.
  • To identify and characterize common errors made by LLMs in classifying these complex queries.

Main Methods:

  • Analysis of 4,161 de-identified Hinglish SRH questions from an underserved community in Mumbai.
  • Annotation of queries into an 8-topic, 40-subtopic hierarchical framework capturing linguistic and cultural nuances.
  • Evaluation of various LLMs (proprietary, multilingual, Indic) using hierarchical F1 (hF1), Exact Match, and topic/subtopic accuracy metrics.

Main Results:

  • Proprietary models, particularly GPT-5 (hF1=0.784), performed best.
  • The Indic LLM Sarvam-M (hF1=0.757) showed strong performance, comparable to leading multilingual models like Claude-3.5-Sonnet and Llama-3.3-70B-Instruct.
  • Models struggled with fine-grained intent recognition, especially with colloquialisms and culturally specific queries, despite generally capturing broad topics.

Conclusions:

  • Hierarchical classification highlights LLM limitations with code-mixed SRH queries.
  • Open-weight Indic LLMs, like Sarvam-M, show significant potential when trained with relevant data and cultural context.
  • Culturally aligned AI tools are essential for improving equitable access to SRH information for underserved global populations.