The project is done as a part of Data Mining project. Instead of using the in-built libraries and interfaces for building a Naive-Bayes data mining algorithm, we use the bare formulae to calculate the actual conditional probabilites and apply the Naive Bayes theorem.
The origin of the data set is from the feedback forms that are submitted by the students of any department of an institution. As of now, there are 2 sets of questions which happens to come from 2 types of forms: Faculty feedback forms and Course Feedback forms.
The input data is first trained by building a matrix of conditional probablilites. Then an overall row-wise and column-wise probablity statistics and then use them to derive two sets of conclusions.
We then use these labelled tuples to output whether the feedback is good or bad