Home
This way you could give an educated guess of what are the most important keywords using a modern approach that is also used by search engines.

It also is more likely to catch the intuitively meaningful words of a text: We all know, that some words are more common than others. While having a hard cut between words that are too common to be included and all the other words is certainly better than nothing, having a smart algorithm decide exactly how common a word is almost certainly even better.

Note: As a side benefit, you don't have to exclude common words (and, I, you, etc.) by hand anymore, because they get weighted down by automatically.

More details? Look here:
https://en.wikipedia.org/wiki/Tf%E2%80%93idf
http://nlp.stanford.edu/IR-book/html/htm......ing-1.html
Author Thumb Christoph shared this idea   06 Nov, 2016

Hide Comments | Hide Activity

Author Thumb Jeffrey commented.   08 Nov, 2016
Interesting idea -- will talk with the programmers and see how feasible this would be to implement.
Author Thumb Jeffrey updated status to Under Review   08 Nov, 2016

Leave a comment...