Made a simple word to emotion classifier using the Naive Bayes Classifier algorithm, with training of 90% and testing of 10%, it provides on average, 51% accuracy based on self-testing. By increasing the size of the dataset, the accuracy and vocabulary will increase, showcased in the vectorisation of the word below, where the one is a recognised word. Mathematically this is represented in a multi-dimensional space based on how many known words the AI understands based on training and testing. Might test image classification next or try new Machine Learning Concepts to create an All-in-one, self taught AI, image, audio, emotion etc.