Create cooccs graph - Githubissues

nilswende / nlp-toolbox

Reimplementation of the Hagen NLPToolbox

2 stars 0 forks source link

Create cooccs graph #4

Closed nilswende closed 3 years ago

nilswende commented 3 years ago

use a sentence file find out its language maybe replace phrases for each sentence:

tag words
phrase removal/stop word removal/stemming
extract the cooccs from cleaned sentence

calculate the association significance for each pair of words

from Cooccs()

nilswende commented 3 years ago

needs to be cut

nilswende commented 3 years ago

cutting done

nilswende commented 3 years ago

the cooccs graph is represented by the DB: total frequency: term.count count of cooccurrences between two terms: count property of their Cooccurs relationship