What it does

This is a script that pulls the most recent 50 tweets from a person’s Twitter account and then uses those tweets to generate a grammar unique to that user.

How we built it

First, our script gathers the most recent 50 tweets from a user’s timeline via the script written by GitHub user yanofsky. We also make use of the Twitter API through the python package, python-twitter. A CSV file is opened in the directory of the script and the tweets are written to that CSV file. Next, we read through the CSV and build a list of tweets, each tweet represented as a string. Those strings are split and then tagged using NLTK’s POS tagger. Finally, we sort the POS tags and build lists of words with the same tag. Analysis of the grammar shows the sorted lists as well as the most used word in each part of speech.

Challenges we ran into

Handling emojis in Python 2.

Accomplishments that we're proud of

Successfully pulled tweets from twitter API and utilized nltk POS tagger and other modules.

What we learned

Learned valuable experience using Twitter API.

What's next for TWIGG: Twitter Grammar Generator


Built With

Share this project: