Let us suppose we have a document where

total_words = 50 (for example -> is,the,now,is,am,here,now) total_unique_words = 40 (for example -> is,the,am,here,now)

How can we apply the Linear Interpolation/ Laplace Smoothening in the case of a trigram

for example

('he') has a count of 12 ('he', 'is') has a count of 8 ('he','is','here') has a count of 4

('is','here') has a count of 6

('is') has a count of 18 ('here') has a count of 5

I think it is related to the count of the values. I have the formula in terms of lambda * P(x_i | x_i-1) but do not know how to implement here. Can someone please give me an PRACTICAL example of how to implement the interpolation

Please make any necessary assumptions in the case as I am learning it the first time