News for hackers. Stuff that matters.
Re:The Amazing Power of Word Vectors (2016) by iconvalleysil on Friday March 03, 2017 @04:04AM (#13781630)
More interesting information from a DOD research lab that resulted in vectors https://www.kaggle.com/c/word2vec-nlp-tutorial/discussion/12...
Re:The Amazing Power of Word Vectors (2016) by charlescearl on Friday March 03, 2017 @07:39AM (#13782448)
Department of Energy, not to be confused with Department of Defense. Granted there is a lot of nuclear weapons work at DOE, but LBNL does mostly big science work up the hill from Berkeley
Re:The Amazing Power of Word Vectors (2016) by j_s on Friday March 03, 2017 @07:42AM (#13782458)
Recently HN featured a podcast interview with "one of the creators of Word2Vec and fastText", documenting some of his professional history.

https://news.ycombinator.com/item?id=13630678

Re:The Amazing Power of Word Vectors (2016) by Bioeye on Friday March 03, 2017 @02:34AM (#13781317)
This is an awesome explanation of those papers! Does anyone have any cool examples of word2vec being used in a project? I'd be interested in seeing what people could make with it.
Re:The Amazing Power of Word Vectors (2016) by mabbo on Friday March 03, 2017 @07:37AM (#13782438)
Document type classification. We wanted to predict which of these k classes a new text document was.

We trained 100-dim word vectors on all the text content we currently have, plus some 30,000 wiki articles related to the business. New content comes in, convert words to vecs, average them, and use that resulting vec as the input to a basic classifier.

For how simple that is, the method is unreasonably good. Widely applicable too.

Re:The Amazing Power of Word Vectors (2016) by ideonexus on Friday March 03, 2017 @10:43AM (#13783586)
For anyone looking for a simple javascript explorable explanation of this you can quickly download and run in a browser, I just found the following GitHub Project.

Demo:

http://turbomaze.github.io/word2vecjson/

Code:

https://github.com/turbomaze/word2vecjson

The code looks pretty straightforward so I look forward to exploring this playground of a new and fascinating concept.

Re:The Amazing Power of Word Vectors (2016) by ideonexus on Friday March 03, 2017 @10:46AM (#13783610)
Update: I'm having a lot of fun exploring the relationships between "nerd," "geek," "dork," etc. in this demo. : )
Re:The Amazing Power of Word Vectors (2016) by vonnik on Friday March 03, 2017 @10:00AM (#13783248)
Word vectors are great. We've also written about them at length.[0] But any one interested in word vectors should also be looking at newer ways of applying neural nets to text. Specifically, convolutional nets with pooling for time are producing great results for clustering and classification.

[0] https://deeplearning4j.org//word2vec

Re:The Amazing Power of Word Vectors (2016) by mustafabisic1 on Friday March 03, 2017 @02:35AM (#13781324)
You nailed it with the "German + airlines" example. Up until that point it was tough to read for a newbie like me. Great blog post