Using Doc2Vec to Suggest SubReddits (opens in new tab)

(reddit2vec.com)

23 pointsjmportilla10y ago12 comments

12 comments

This is pretty neat, but the biggest problem for me is the case sensitivity; reddit itself doesn't use case sensitivity, so it's hard to remember the exact capitalization of a subreddit name.

jmportillaOP10y ago

Yeah, I know its super annoying. Hopefully I'll have time to update the model with lowercase names sometime next week.

utunga10y ago

Hi!

Great work. I guess my question is - do you use 'averaging' of word vectors or the Chinese Restaurant process - to get to sub reddit vectors. You describe the Chinese Restaurant process as a "more sophisticated method" that you "can" use, but in my experiments with word2vec and reddit (https://github.com/utunga/gensimred) I quickly discovered that simple averaging just does not work. Averaging has this awful 'revert to mean' thing that turns all the paragraph vectors into a sort of bland gray goo where they are all the same.

If you did use Chinese Restaurant process (I love that phrase - brings back memories of an occasion at a Dim Sum restaurant where this almost literally happened) it'd be great to see any source code you may feel like releasing ;_) ... well, it can't hurt to ask..

gojomo10y ago

The gensim library now offers the 'Paragraph Vector' [1] approach to create vectors for longer ranges of text. It's available in a class named Doc2Vec – but I don't think that's what is being used here.

The Paragraph Vector approach can give interesting results for document-similarity, including similarity after certain 'algebraic'-like additions/subtractions of other topics/word-concepts. [2]

[1] http://arxiv.org/abs/1405.4053

[2] http://arxiv.org/abs/1507.07998

jmportillaOP10y ago

I used the gensim Doc2Vec implementation. You can check out some of the source code here: https://github.com/jmportilla/Reddit2Vec

utunga10y ago

Hi... Thanks for that. Awesome and much appreciated.

joelthelion10y ago

Very cool. Little tip: use "-funny" to get high-quality subs :)

Yadi10y ago

Awesome seeing someone use the reddit dataset :)!

Wouldn't a w2v as a recommender for the user might have been better?

Taking user's comments/likes/subreddits as a feature.

jmportillaOP10y ago

I think your thinking of just a classic collaborative filtering recommendation system.A simple w2v system would take into account all words, then have to be filtered by words that are equal to subreddits. Although, I may have misunderstood your suggestion.

riffraff10y ago

neat, I'd suggest considering spaces as "+" i.e. "cats awww" should be the same as "cats+awww" I guess :)

haxiomic10y ago

Nice idea :), works well. Spotted a small typo in the examples:

pcmasterace+mac should be pcmasterrace+mac (missing an r)

jmportillaOP10y ago

Thanks! I'll fix it

j / k navigate · click thread line to collapse

12 comments

sdrothrock10y ago

This is pretty neat, but the biggest problem for me is the case sensitivity; reddit itself doesn't use case sensitivity, so it's hard to remember the exact capitalization of a subreddit name.

jmportillaOP10y ago

Yeah, I know its super annoying. Hopefully I'll have time to update the model with lowercase names sometime next week.

utunga10y ago

Hi!

gojomo10y ago

The Paragraph Vector approach can give interesting results for document-similarity, including similarity after certain 'algebraic'-like additions/subtractions of other topics/word-concepts. [2]

[1] http://arxiv.org/abs/1405.4053

[2] http://arxiv.org/abs/1507.07998

jmportillaOP10y ago

I used the gensim Doc2Vec implementation. You can check out some of the source code here: https://github.com/jmportilla/Reddit2Vec

utunga10y ago

Hi... Thanks for that. Awesome and much appreciated.

joelthelion10y ago

Very cool. Little tip: use "-funny" to get high-quality subs :)

Yadi10y ago

Awesome seeing someone use the reddit dataset :)!

Wouldn't a w2v as a recommender for the user might have been better?

Taking user's comments/likes/subreddits as a feature.