[Source code and documentation sharing] CRF ++ based on the Chinese named entity recognition program

1 experimental environment

• win10

• Python 3.7.0

• CRF ++ - 00:58

2 main program algorithm

Conditional Random Fields (CRF):

Condition with a defined airport: Let G = (V, E) represent an undirected graph, any node v corresponding to a random variable Y, and therefore Y = (v ∈ V), the apex portions Y elements and the undirected graph G a a correspondence. When at Χ, the conditional probability distribution of the random variable Y v FIG subject Markov property: p (| Χ,, w ≠ v) = p (| Χ,, w ~ v), wherein w ~ v represents ( w, v) is the edge of the undirected graph G. Then we call (Χ, Y) is a condition with the airport.

 

Reference documentation and complete documentation and source code Download:

https://www.write-bug.com/article/2356.html