The Dataset
This is the set of queries that were hand constructed keeping in mind that the queries could obtain answers from U.S. Census Bureau conducted Annual Retail Trade Survey of U.S. Retail and Food Services Firms for the period of 1992 to 2013 (link).
There are 293 queries in all and a few sample queries have been pasted below
This dataset was used in the paper (Adapting general-purpose speech recognition engine output for domain-specific natural language question answering) presented during the IJCAI 2015 Workshop on "Replicability and Reproducibility in Natural Language Processing: Adaptive methods, resources and software" in Buenos Aires, Argentina.
How did we Generate the dataset?
We looked at the data (link) specifically for the duration 1992 to 2013 and hand constructed the text queries. A set of 6 people (L2 English) generated 50 text queries (300 queries in all). They were asked to construct queries so that an answer from the data at link could be obtained. Then we removed duplicate queries; this left us with 293 queries in all.
How did we use it?
The slides should be of use to understand how the data was used.
Can I use it?
You could download the attached xls (openoffice format!) and you could use the queries (in ASCII).
However a line of acknowledge will be a good gesture ;-)
@misc{queries_dataset,
author={AwazYP},
howpublished={\url{https://sites.google.com/site/awazyp/data/ijcai}},
note={Handcrafted at TCS Innovations Lab - Mumbai, Delhi},
year={2015},
}
I have a Query?
Send an email: x.y@company.com (x:sunilkumar; y:kopparapu; company:tcs|gmail)