This folder contains the data used in experiments for the paper.

It contains processed data for the following publicly available datasets:

1) Incar : https://nlp.stanford.edu/blog/a-new-multi-turn-multi-domain-task-oriented-dialogue-dataset/

2) CamRest : https://www.repository.cam.ac.uk/handle/1810/260970

3) Maluuba Frames : https://datasets.maluuba.com/Frames

Each sub-folder contains the train, validation and test split of the dataset.
Each data sample contains the following:

1) Dialog context
2) Gold response
3) KB queries (if any)
4) KB results corresponding to the queries.

Before running the code, you will need to download the Glove vectors into the "data" folder.
Download the 200 dimensional vectors and name the file: glove.6B.200d.txt