This repository contains a dataset for training a mental health chatbot using natural language processing techniques. The dataset includes a collection of user messages labeled with their corresponding intents and entities. Intent Categories
Greeting: User initiates conversation with a greeting message
Symptom Query: User expresses concerns about a specific symptom
Treatment Query: User asks about treatment options or recommendations
Crisis: User expresses suicidal thoughts or intentions
Gratitude: User expresses gratitude for the chatbot's assistance
The dataset also includes entity recognition labels for specific symptoms and treatments. The following entities are included in the dataset:
Depression
Anxiety
Panic Attacks
Cognitive Behavioral Therapy
Medication
The dataset is in a CSV format with two columns: message and label. The message column contains the user message and the label column contains the intent or entity label. Dataset Size
The dataset contains a total of 10,000 labeled user messages. The messages were collected from various online mental health forums and support groups. Data Preprocessing
The dataset has been preprocessed to remove personally identifiable information and other sensitive data. The messages have also been cleaned and tokenized to prepare the data for machine learning model training. License
This dataset is released under the MIT License.
We hope that this dataset will be helpful for researchers and developers interested in building mental health chatbots using natural language processing techniques. Please feel free to use this dataset for academic or commercial purposes and let us know if you have any feedback or suggestions for improvement.