This repository various small scale projects based on Natural Language Processing.
Some of the most significant ones are

1. Part of Speech Tagger
2. Password Strength Classifier
3. Topic Modeling using Non-negative Matrix Factorization

We incorporate context of a word in our model to assign it a Part of Speech
Eg. The word plant may be a verb or a noun depending on the context

Let's plant a tree. --> plant is a Verb here
Photosynthesis is a transfer of energy from the Sun to a plant. --> plant is a Noun here

It is a supervised learning based model which can be used to classify a password as a strong, weak or medium
Used Tfidf Vectorizer from sklearn and a Tokenizer which splits a password in its constituent characters
Got 98.63% accuracy on test data and 99.10% accuracy on training data using XGBClassifier from XGBoost

Topic Modeling is an unsupervised learning algorithm which enables one club a group of articles, books, tweets etc into different topics to which they belong. For eg. Politics, Wellbeing, Sports, Cinema, Culture, Climate, etc