Amfam chatbot intent dataset(ACID) contains 174 unique intents related to topics driven by customers contacting an insurance company. Each intent represents a particular course of action for the chatbot. The intent prediction dataset was collected from past interaction of customers with our service representatives at American Family Insurance. Subject matter experts created different intent categories and handpicked examples belonging to each intent category. The dataset contains 175 unique intents. It is split into a training set that contains a total of 11,130 examples and test set that contains total of 11,042 examples. The distribution across the intent classes is highly skewed with the smallest intent class containing 10 examples and the biggest class containing 378 examples in the training set.
Checkout our paper http://ceur-ws.org/Vol-2666/KDD_Converse20_paper_10.pdf
If you find the dataset helpful in your paper, please site our work
Acharya, Shailesh and Fung, Glenn: Using Optimal Embeddings to Learn New Intents with Few Examples: An Application in the Insurance Domain. Proc. KDD 2020 Workshop on Conversational Systems Towards Mainstream Adoption(KDD Converse 2020), 2020, CEUR-WS.org, online CEUR-WS.org/Vol-2666.
Bibtext
@article{acharya2020using,
title={Using Optimal Embeddings to Learn New Intents with Few Examples: An Application in the Insurance Domain},
author={Acharya, Shailesh and Fung, Glenn},
year={2020},
conference={KDD 2020 Workshop on Conversational Systems Towards Mainstream Adoption(KDD Converse 2020)},
publisher={CEUR-WS.org}
url={http://ceur-ws.org/Vol-2666/KDD_Converse20_paper_10.pdf}
}