Algorithm project

my project in hybridizing naïve Bayes classifiers (baseline + multinomial + Bernoulli + semi supervised ) to be one classifier one algorithm for twitter sentiment analysis. but the last freelancer said we can not hybridize MN+BN, so he did (baseline + mnb + ssnb) and (baseline + bnb + ssnb) my first question, can we hybridize (baseline naïve bayes + multinomial naïve bayes + Bernoulli naïve bayes + semisupervised naïve bayes ) to be one classifier one algorithm with high classification accuracy using two features are (time and location from twitter) if your answer is yes, so can you do it?.......... if your answer no, I should explain what I need from you to update my available program now