How can I get the similarity degree of a recognizing result compared to the frequent words list ?

[葉家忠]

Hello, 

I tested the Tesseract that  the accuracy will be highly increased by using frequent words list which is added the the traineddata as the freq-dawg form. 

The originally error result will be fixed according to the  frequent words list to become correct one. (it's magic!!)

Now I want to know if a word if is recognized, then how the Tesseract judge if the result is among the frequent words list? (which code snippet can I study?)

Tesseract will fix the result if the original result is similar to one word from the  frequent words list. 

How can I get the similarity degree if the result is fixed to become one of the the word from frequent words list? 

(the similarity degree between the recognized word and one of the frequent words list must be high enough to make such a decision)

What I want is to get the similarity degree (or matching rate?) from the recognizing process. (I want to fixed the source code to output the similarity degree)

Thanks in advance.

(Sorry for my poor English, if you don't catch my problem , please let me know ^^)