This is a repository for all projects within the natural language processing class
Created Naive Bayes and logic regression classifiers for types of businesses and comparing the classifiers' results Classifiers are generated from yelp reviews of businesses and attributes of businesses.
Utilized unigram, bigram, trigram, and smoothed n-gram to predict most likely choices of blanked words for a given sentence. Dataset used to test the models came from Microsoft's MSChallenge
Created set of regular expression rules without use of library packages to convert numbers to texts. Example: "The total costs $1.50" --> "The total costs one dollar fifty cents"