search

kossiitkgp / kwoc-2018

Web app for Kharagpur Winter of Code, 2018
https://kwoc.kossiitkgp.org
MIT License
0 stars 0 forks source link

Run Tag Scraper #93

Closed thealphadollar closed 5 years ago

thealphadollar commented 6 years ago

Once we've finalized the projects we can run tag scrapper to ensure that the projects also have the tags they are associated with on github along with the ones set by the mentor manually.

To be done before projects go live.

Ayushk4 commented 6 years ago

I think that this may not be the best possible way to get the projects tagged. The response sheet isn't storing the tags for a repo. And, for some tagged projects, since this is being centered (not limited though) in Kharagpur, people might wanna add extra tags. Take for example Kronos and some other metakgp project, where we might wanna add the tag - academics-kgp or something so that the student know that the project is centered around Kharagpur. Letting mentors do the tagging, will make sure that the list of tags is exhaustive.

I think it's better if we requested mentors to give add tags to their projects once selected. A lot of repos don't usually have tags, or have incomplete number of tags. Its not good to run the tag scrapper as well as get it done manually by mentors, so we should just ask mentors to do it.

Hence please close this issue, or keep it for next year, the script for tag scrapping should be automatically called once the mentor fills the link to the project in the mentor form, while keeping the option to manually add more tags. So keep this issue and modify it low priority for next yr.

Ping @thealphadollar .

themousepotato commented 6 years ago

I disagree with @Ayushk4. Most of the projects don’t have any tags associated with them. We need to do it manually or ask mentors to do it after running this.

On Thursday, November 29, 2018, Ayush Kaushal notifications@github.com wrote:

I think that this is unnecessary.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/kossiitkgp/kwoc/issues/93#issuecomment-442591950, or mute the thread https://github.com/notifications/unsubscribe-auth/AgXlPh4u8XkzgcIb7RfpYr-SXLLAOmFNks5uzvD0gaJpZM4Y4HhC .

arnav-t commented 5 years ago

I think we can do the broad tagging (like for languages) automatically (perhaps the same way Github identifies languages?). We can do the more "nuanced" tags manually and leave the option for further tagging open to the mentors.
This way, we can ensure we have working and uniform tags for all our projects while giving the mentors freedom to promote their own projects.

thealphadollar commented 5 years ago

Well, we can do that as well. @TheLethalCode Can you please do that? The script is present in gh_scripts.

themousepotato commented 5 years ago

@Ayushk4 In case you're interested, https://medium.com/coinmonks/multi-label-classification-blog-tags-prediction-using-nlp-b0b5ee6686fc. Input: README :P