aggies 19.0%
Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
11–20 of 82 posts
Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#12Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#13You should probably add how exactly you retrieved the 4 million Reddit posts.
Probably with the official Reddit API? There are several libraries for it.
It took me 36 hours to collect the 4M posts. Reddit API returns results in batches of 100 results, and then sleeps for 2 seconds.
You can find some more details on how it was built here https://blog.valohai.com/machine-learning-pipeline-classifyi...
I can publish on Github the repository that runs two commands to collect the data if your are interested.
Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#14Would be nice to have the subreddits be links. So I could just click it to open a new tab of that subreddit.
Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#15You should probably add how exactly you retrieved the 4 million Reddit posts.
Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#16Anyone remember /r/reddit.com? That was around the time reddit looked like Hacker News and people were embarrassed to admit they use it.
Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#17What are some examples where the model does recommend meaningful things?
Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#18Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#19Earlier quoted context omitted.
Probably with the official Reddit API? There are several libraries for it.
Right! I used the official Reddit API. I created an APP, got the credentials for the API. Then used the Python library PRAW to consume the API. https://praw.readthedocs.io/en/latest/ It took me 36 hours to collect the 4M posts. Reddit API returns results in batches of 100 results, and then sleeps for 2 seconds. You can find some more details on how it was built here https://blog.valohai.com/machine-learning-pipeline-…
I was curious since it seems like it was using the BigQuery dataset, but PRAW works too.
Re: Show HN: Subreddit Finder - Trained on 4M Reddit Posts from 4K Subreddits
#20Anyone remember /r/reddit.com? That was around the time reddit looked like Hacker News and people were embarrassed to admit they use it.
It's not just HN's aesthetic that is minimal and no-nonsense, it's their moderation policies and the tone they set for the community. There is perfect alignment between their approach to content, community, and UX—no fluff, no nonsense, no manipulation, just the simplest, most valuable material possible.
If they expand the scope of acceptable content, it will be really hard not to tweak moderation policies, and eventually you end up with something like reddit, each subreddit might as well be its own (typically under-staffed) site.
EDIT: This is not critical of any post on this thread, just seeing the above comment about old-school reddit got me thinking.