Other Question

We want to build a naive Bayes sentiment classifier using add-1 smoothing. Here is our training set: 

   – the movie has no plot 

   – honestly pretty boring 

   + pretty interesting movie 

Test Set:  

   pretty enjoyable plot 

Answer the questions below given the sets above: 

Compute the prior for the two classes + and -, and the likelihoods for each word given the class (leave in the form of fractions). 

Would using binary multinomial Naive Bayes change anything? 

1. Why do you add |V| to the denominator of add-1 smoothing, instead of just counting the words in one class? 
2. Watch the following videos in the lecture series to prepare you for this Assignment:
3. 3.1 Intoduction to N-grams (time 8:42)

3.5 Smoothing: Add-One (time 6:31)

https://www.youtube.com/playlist?list=PLoROMvodv4r…