[–]moridin007 1 point 1 day ago
comparing this algorithm with the one i wrote for wingztv.com (shameless plug) i am really shocked by how much better these guys algo is.
although mine was just kinda was hacked up. it just sees sentences with more words in common and ranks then by it, and then i just grab the top 5. this algo looks a lot more fancy O.o probably why it's a lot better lol. http://www.reddit.com/r/InternetIsBeautiful/comments/2wqq5t/...
From a simple test, this tool is at the hobby level. The text is not interpreted semantically, because the so-called summarization is nothing but blocks of text from the original text, copied ad-literam. It lacks any kind of rewording. My guess is that for content retrieval it employs the the most rudimentary model one can think of [1], does a plain token indexation, then chooses the blocks with the highest density of non-ordinary words.
This is no hobby level system. I've done side-by-side comparisons and they've got something else going on here that beats other systems hands down. You've also seemed to miss the fact that these guys are enabling the context to be controlled using any word.
11 comments
[ 1.2 ms ] story [ 43.5 ms ] threadFrom a simple test, this tool is at the hobby level. The text is not interpreted semantically, because the so-called summarization is nothing but blocks of text from the original text, copied ad-literam. It lacks any kind of rewording. My guess is that for content retrieval it employs the the most rudimentary model one can think of [1], does a plain token indexation, then chooses the blocks with the highest density of non-ordinary words.
[1] http://en.wikipedia.org/wiki/Standard_Boolean_model
This system utilizes vector space and is based on a development at Lawrence Berkeley National Laboratory:
https://www.kaggle.com/c/word2vec-nlp-tutorial/forums/t/1234...
http://genopharmix.com/biomimetic-cognition/in_silico_cognit...
It's designed to mimic the process of human cognition aka AI.
It's fairly sophisticated as shown by its results.