support: Sample Dataset in Installation Documentation

[adm-2k]

Support Question

Can there be a sample dataset linked to in the documentation and/or hosted in the repository, so a new user can easily test out the tool? Preferably, one of the smaller datasets with a high signal of CIB.

[adm-2k]

Here is a link to the current Testing documentation page; however, all of the links in the example README are broken.

[KristijanArmeni]

Thanks for opening this @adm-2k ! Indeed, we need a good, small but preferably real-world demo dataset that can be either a) shipped with the app or b) users can easily obtained it form some publicly-hosted place.

This is related to issue #226 -- I've shared some initial ideas in my #226 (comment) before. Copy-pasting the main part:

I think the ideal way to do this would be:

1. take a real social media dataset with bot content
2. Truncate it to fit the [file size limits on GitHub](https://docs.github.com/en/repositories/working-with-files/managing-large-> files/about-large-files-on-github), < 50MB. I think that should still allow us to have a .csv/.parquet dataset with ~10k rows; which is sufficient for playground
3. The dataset should have the required columns for the copy-pasta and the hashtag tests (the two tests that have dashboards currently) such that users can get started without too much friction.

I think the Russian Trolls dataset would be a good candidate.

Currently nobody as actively working on this, but I think it would be great to have this in for the next release (v0.11.0). Any interest in picking this up @adm-2k ? Happy to provide guidance and think along.

[KristijanArmeni]

Here is a link to the current Testing documentation page; however, all of the links in the example README are broken.

The testing documentation page you shared refers to testing the code for development when preparing new contributions (i.e. testing that new contributions are not introducing breaking changes in code), but not testing the app for the users. The tests for code have their own very tiny synthetic datasets. In my opinion, the best place for user-testing instructions would be in the user manual that Ben is developing. And for that, we currently don't have a good ready-to-use dataset indeed.

I clicked the links in the readme and they worked for me (led me to relevant code snippets)? (unless I'm missing something!)