Skip to content

Enquiry into simple & complex table count ratio #23

Description

@mckhang

Hi, thanks for your great work.

I would like to raise an enquiry regarding the count of simple and complex tables in SynthTabNet dataset.
In Table 2 of the paper, there is a comparison of TEDS score for Simple, Complex and All.
Upon downloading the dataset and checking for the number of simple and complex tables in test dataset,
taking the definition of 'complex' as 'whether there is rowspan or colspan' in the HTML seq, the ratio of
simple : complex turns out to be 0.2 : 99.8.

Is this the expected ratio of simple to complex tables? Given this ratio, the result of TEDS score under 'All' column
does not match that of 'Simple' and 'Complex' as shown below:
(0.002 * 96.9) + (0.998 * 95.7) = 95.7 which is not equals to the reported 96.7

Hope to seek your clarification in this matter. Thank you

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions