Recently in learning, text classification with CHI do feature extraction, met a question want to ask next everybody, see below
Figure table 1 is very common for an example of a chi-square figure, according to table 1 I had table 2 and table 3 and table 2 and table 3 according to chi-square formulas of value should be the same, but the word "apple" clearly do not belong to the sports! But the word "apple" score is like "basketball", will be extracted as features, is this why? My understanding of the characteristics of the problem? Trouble you help me to solve, thank you!!!!!

CodePudding user response:

Before square, basketball is positive, the apple is negative, but after a square is the same

CodePudding user response:

Although apple is not the characteristics of sports, but are the characteristics of the sports