shuzi / insuranceqa Goto Github PK

I think it would be beneficial to include an example of the data in the ReadMe.

I opened the files but it's a bit difficult to understand what is written. Could you provide an example in english and not with tokens?

In the README.md, Pool is mentioned at Pool is generated by SOLR.
I haven't check the data yet, but based on my understanding, pool data is retrieved with Solr system and it is the questions array list when query with the question in the first column. Is it?

<Domain><TAB><QUESTION><TAB><Groundtruth>

So, Groundtruth is the answer's label. Pool is the similar questions that return from Solr.

What would be the license for this data set? Would it be possible to be used for commercial purposes as well?

I know that the Readme says,

This dataset is provided as is and for research purpose only

But wanted to make sure and confirm. Thank you.

Are variations with punctuation of a certain word (e.g. "titles.", "titles,") intentional? It seems like there is a lot of redundancy.

I have check some papers, but they all experiments on version 1 dataset, is there any experiments on version 2 dataset?