Authors: Olumide Ebenezer Ojo, Olaronke Oluwayemisi Adebanji, Alexander Gelbukh, Hiram Calvo, Anna Feldman

Published on: February 06, 2024

Impact Score: 8.12

Arxiv code: Arxiv:2402.04442

Summary

• What is new: This research explores the use of advanced text embeddings and one-shot classification systems in healthcare communication, showcasing how AI can distinguish between doctor-written and AI-generated texts.
• Why this is important: The challenge of effective communication between healthcare providers and patients, specifically in classifying different sources of medical consultation texts.
• What the research proposes: Utilization of state-of-the-art embeddings and one-shot classification systems to analyze and classify healthcare consultation texts.
• Results: Embeddings such as Word2Vec, GloVe, Character n-grams, and GPT2 were effective in capturing semantic features, with machine learning architectures enhancing the quality of health conversations.

Technical Details

Technological frameworks used: Various embeddings (Word2Vec, GloVe, fastText, GPT2) and one-shot classification systems

Models used: Bag-of-words, character n-grams, Word2Vec, GloVe, fastText, GPT2 embeddings

Data used: Healthcare consultations text data

Potential Impact

This research could significantly impact the healthcare industry, particularly in telemedicine platforms, health informatics companies, and AI-driven healthcare solutions.

Want to implement this idea in a business?

We have generated a startup concept here: MediComm AI.