ads_tweetapp's Introduction

Analyze Twitter Data for Sentiment Analysis

Flask web app for Twitter data analysis of Taco Bell food chain tweets

Code Documents

W2V.pkl : The Word2Vec model, saved in full, for access in the wordcloud function.
sent_df and tokens JSON files save the work done in the preproc.ipynb notebook, for easier access to the processed data.
tweets.db : The sqlite database that stored the accumulating tweets run by the getTweets.py script.

This is the main Flask app directory.
The db.py script defines the SQLite database.
mkclouds.py is a script that generates the wordcloud images from the similar words as found by the word2Vec model.
word_vectors.kv is the stored KeyedVectors from the Word2Vec model.
HTML models are in the templates directory.

Procfile is a simple script to allow deployment to Heroku.
getTweets.py is the Python script used to cull tweets via the Twitter API v 2.x.
The preproc Jupyter notebook is a place for experimentation and general pre-processing of data with Pandas. The Word2Vec modeling and WordCloud functions were developed here.
The requirements.txt is for deployment, and lists the installed packages in the pip virtual environment.
tweetTasks.ipynb is the initial work on developing the getTweets script and other code development for processing the data.
twitDBTest.py was written to test whether the ongoing crontab job was successful, and how many tweets were appended to the database.

Recommend Projects