$($a$)$ The following text sentences were taken from various reviews of products together with whether the review was positive or negative.

$\backslash$table$[[$Review text,tag$],[\text{'}$A great product, must buy',positive$],[\text{'}$Different than the description',negative$],[\text{'}$The product delivered quickly',positive$],[\text{'}$Good condition',positive$],[\text{'}$The product stopped working quickly',negative$],[\text{'}$The product was faulty',negative$]]$

Suppose we are given a review with text 'Must buy quickly, inaccurate description', but we are not told whether the review was positive or negative. We would like to use a Naive Bayes classifier to decide whether to record the review as positive or negative.

i$.$ What would be the recommendation of the Naive Bayes classifier using maximum likelihood? Provide a justification for your answer.

ii$.$ Repeat part i$.,$ using Laplace smoothing with $=1$

iii. Suggest two potential ways to improve predictive performance.

NOTE$-$ Plese provide handwritten answers as they are easy to understand.