Healthcare Chatbot Intent Dataset

This dataset provides detailed, synthetic healthcare chatbot conversations with annotated intent labels, message sequencing, and extracted entities. Designed for training and evaluating conversational AI, it supports intent classification, dialogue modeling, and entity recognition in healthcare virtual assistants. The dataset enables robust analysis of user-bot interactions for improved patient engagement and automation.

  • opened 12 times
  • last updated 29 Jul 2025
  • by GoMask
The brief that made it

Training and benchmarking intent classification models for healthcare chatbots

Sample rows

preview · 8 of 200 rows · all 11 columns
transcript_idstringconversation_topicstringconfidence_scorefloatintent_labelstringmessage_idstringmessage_orderintegersender_typestringmessage_textstringtimestampdatetimeentities_extractedstringuser_idstring
T001appointments0.94book_appointmentM00011userHi, I need to book an appointment with Dr. Smith.2021-04-05T09:32:15Zappointment_date:2021-04-12U1023
T001appointments0.97request_timeM00022botWhat time would you prefer for your appointment?2021-04-05T09:32:20Zblankblank
T001appointments0.91provide_timeM00033user10:30 AM works for me.2021-04-05T09:32:27Ztime:10:30U1023
T002symptoms0.88symptom_checkM00041userI have a headache and a sore throat.2022-02-15T13:05:51Zsymptom:headache,symptom:sore throatU1045
T002symptoms0.86request_durationM00052botHow long have you been experiencing these symptoms?2022-02-15T13:06:05Zblankblank
T002symptoms0.84provide_durationM00063userAbout three days.2022-02-15T13:06:13Zduration:3 daysU1045
T003medication0.81location_queryM00071userCan you tell me where the nearest pharmacy is?2023-10-11T17:14:55Zlocation:pharmacyU1098
T004insurance0.91insurance_updateM00081userI want to update my insurance information.2022-09-23T08:11:34Zold_insurance_number:JS9182,new_insurance_number:SD2735U1122

What the 200 rows show

from the 200-row sample

Other (conversation topic) stands out: mean confidence_score is 0.49, against 0.82 for the rest.

  • 0.82median confidence_score
  • 2sender types
  • 1median message_order
Mean confidence_score by conversation_topic200 rows
00.510.49other45 ro…0.90appo…40 ro…0.86symp…39 ro…0.76medi…26 ro…0.69medi…18 ro…0.81insu…16 ro…0.75vacc…16 ro…
confidence_score200 rows, in bands of 0.1
040804031213111326764200.51confidence_score →

Median 0.82, from 0.0 to 0.98.

intent_label200 rows · top 10 of 17 values
  1. other32
  2. symptom_check20
  3. greeting16
  4. prescription_issue15
  5. insurance_update14
  6. book_appointment12
  7. request_duration11
  8. language_request11
  9. allergy_query11
  10. vaccine_eligibility11
11 columns by typefrom the column list below
  • string 8
  • integer 1
  • float 1
  • datetime 1

Columns

11 columns in three groups
blueprint · 11 columns
columntypedescriptionexample
Text 8 columns
transcript_idstringUnique identifier for each chat transcript.uniqueT001
message_idstringUnique identifier for each message within a transcript.uniqueM0001
sender_typestringIndicates whether the sender is the user or the bot.user · botuser
message_textstringThe actual text content of the message.10:30 AM works for me.
intent_labelstringAnnotated intent label for the message, such as 'book_appointment', 'symptom_check', etc.book_appointment
entities_extractedstringComma-separated list of key entities extracted from the message (e.g., medication names, symptoms, dates).optionaltime:10:30
conversation_topicstringHigh-level topic or category of the conversation (e.g., 'appointments', 'medication', 'insurance').7 topics · optionalappointments
user_idstringAnonymized identifier for the user participating in the chat.optionalU1023
Numbers 2 columns
message_orderintegerOrder of the message within the transcript, starting from 1.1 or more1
confidence_scorefloatConfidence score (0-1) for the assigned intent label.0 to 1 · optional0.94
Dates and times 1 column
timestampdatetimeTimestamp indicating when the message was sent.optional2021-04-05T09:32:15Z

15 SQL practice questions for the Healthcare Chatbot Intent Dataset

Written for this data and checked against it: 12 SQL and 3 Power BI exercises from beginner to advanced. Try the first three free; the practice kit has all 15, a starter .sql file and an answer key with every expected result.

  1. 01BeginnerFreeSELECT · WHERE · ORDER BY

    User messages about medication

    Return message_id, transcript_id and message_text for every message where sender_type is 'user' and conversation_topic is 'medication'. Order the results by message_id ascending.

    Tables: healthcare_chatbot_intent_dataset · your result: 3 columns, 26 rows
  2. 02BeginnerFreeDISTINCT · ORDER BY

    Distinct intent labels

    Return each distinct intent_label found in the data, sorted alphabetically. The result should have one column: intent_label.

    Tables: healthcare_chatbot_intent_dataset · your result: 1 columns, 17 rows
  3. 03BeginnerFreeGROUP BY · COUNT

    Messages per topic

    Count the messages for each conversation_topic. Return conversation_topic and message_count, ordered by message_count descending and then by conversation_topic ascending.

    Tables: healthcare_chatbot_intent_dataset · your result: 2 columns, 7 rows
  1. 04Average confidence by senderBeginner
  2. 05Power BI: user messages by topicPower BI
  3. 06Intents with many messagesIntermediate
  4. 07Confidence bucketsIntermediate
  5. 08Messages per yearIntermediate
  6. 09Topics below overall confidenceIntermediate
  7. 10Power BI: messages and transcripts by orderPower BI
  8. 11Top intent per topic by confidenceAdvanced
  9. 12Seconds between messagesAdvanced
  10. 13Running message total by monthAdvanced
  11. 14Topic share of messagesAdvanced
  12. 15Power BI: low-confidence intent anomaliesPower BI

Get the practice kit: the full dataset, all 15 exercises, a starter .sql and the answer key. One-time $3.99.

Get the practice kit · $3.99

Use it for

  • median confide…0.82200 rowsmean confidence score…0.49other0.90appo…0.86symp…0.76medi…

    A healthcare dashboard

    Confidence_score by conversation_topic and a breakdown of intent_label. Excel, Power BI or Tableau.

  • Why do the 45 other rows have a mean confidence_score of 0.49?

    A root-cause class exercise

    Hand out the rows and one question. The answer is in the data, not in the brief.

  • A software demo

    Believable transcripts with message_id, message_order and sender_type to fill a screen in front of a buyer.

Not quite right?

Make it yours.

Same 11 columns, your size and your rules. See 20 rows before you pay.

Preview 20 rows free

1,000 rows: $1.99 · 10,000: $12.99One-time. No subscription. All prices

This dataset200 rows11 columns
Yours1,000 rows11 columns

blueprint · healthcare-chatbot-intent-dataset

Behind this dataset

Same schema. As many rows as you need.

These 200 rows came out of a blueprint — 11 columns with generation rules behind each one. Open it in Data Factory to retune a column, add your own, wire in foreign keys, and run it at the size you actually need.

Rules it was built with
  • Each message tagged with one intent
  • Multiple intents possible per session
  • Sensitive information removed
  • Session lengths vary by user type
Rows
Open the blueprint in Data Factory

1 credit per row. Preview 20 rows free. Buy rows for the full run.

Exports
CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
Licence
yours to use, including commercially
API slug
healthcare-chatbot-intent-dataset

More in Healthcare

Browse all
Also in HealthcareTelemedicine Session Audio TranscriptsEmergency Department Triage PatternsPatient Vital Sign Time SeriesAll 1,966 datasets
In collectionsSentiment and Text Classification Datasets

What should your data show?

Preview 20 rows free
No signup. No card.