python - Naive Bayes, dataset choice(sentences vs dictionary) -
i'm trying classify emotion based on text using naive bayes. have isear dataset , nrc dataset. felt isear has lower result compared nrc. little explanation didn't know difference between isear , nrc, isear dataset consist of sentences , nrc word dictionary. result far expected when inputting manual sentences using isear.
i'm kinda new machine learning, correct me if i'm wrong.
so naive bayes worked using prob of each word showing right? example, have word "i'm happy" , appears on "joy" features 5 times , 6 times on "surprise" features. isn't cause false predict? compare word dictionary, example, happy labeled joy , surprise , occur once in each dataset?
am okay go if using word dictionary data set using simple naive bayes method?
Comments
Post a Comment