A web application using NLP to create definitions for slang words based on context from tweets
This application employs machine learning and natural language processing to evaluate the definition of slang words based on the context of the tweets they are used in. We pull tweets from the Twitter API that contain the word, and add them to our data set of sentences, some of which do and others that do not include the word. With this data set, we create a word vector model that maps out all of the words to representative vectors based on their relation to other words so that they are easier to compare. From there, we do a vector comparison that finds the most similar vectors to that of the input slang. We extract the words that the vectors represent to find synonyms of the input word, which is then returned to the user on the front end.
dude relates to friend, friends, or man fam relates to dude, friends, or friend bro relates to cool, please, or man bruh relates to man, also, or please cool relates to awesome, follow, or best bro relates to see, os, or friends
To install dependencies, run:
$ pip install -r requirements.txt
To run the app, use the following command:
$ python main.py
By default, it will run on