MedDialog English Doctor-Patient Conversations

MedDialog contains 0.26 million English-language dialogue exchanges between doctors and patients. The dataset is sourced from healthcaremagic.com and icliniq.com and is maintained by author bigbio, with a last recorded update in December 2022.

Use Cases

• Train a dialogue generation model on the doctor-patient conversation structure to simulate medical consultations.
• Analyze conversation patterns and topics within the large-scale text corpus of medical dialogues.
• Extract and classify medical intents or advice from the English text of patient queries and doctor responses.
• Build a retrieval-augmented model for healthcare Q&A using the dialogue text as a knowledge source.

Strengths

• Contains 0.26 million dialogues, providing a large-scale text corpus for analysis.
• Sourced from two established medical Q&A platforms, healthcaremagic.com and icliniq.com.
• Specifically focuses on the structured domain of English doctor-patient conversations.

Limitations

• Lacks documented column structure or sample data, hindering immediate understanding of data format.
• Potential bias towards the style and scope of consultations presented on the two source websites.
• Update recency is unknown beyond the last platform-recorded update in 2022.

Provenance

Source

healthcaremagic.com and icliniq.com

Collection Method

Conversations scraped or collected from medical Q&A websites.

Time Range

null

Freshness

null

Geography

null