ThinkOpenly

---

Demo: https://www.youtube.com/watch?v=z9iJFXSoQCM

How to run

1. After cloning repo make sure you have the latest nodejs LTS and python3.8 installed
2. Install pipenv

cd backend
python -m pip install pipenv

3. Install deps

pipenv install

4. Set up django

cd config
cp .env.sample .env
cd ..
pipenv run python manage.py migrate
pipenv run python manage.py runserver

5. Open another terminal window to run the frontend and run the following commands

npm install -g yarn
cd frontend
yarn
yarn start

Note: You must have both the backend and frontend running, specifically on ports 8000 and 3000 respectively.

Inspiration

We were inspired to do this project in light of the women’s advocacy and their troubles. One of our members was scrolling through instagram and saw this and thought that it would be useful to have a similar platform catered to individuals to post about their political views where policy makers can actually see support for certain topics.

What it does

ThinkOpenly is similar to StackOverflow or StackExchange. Users can post their ideas on certain political topics and the aim is to have real world politicians be able to respond and view the voices of who they represent (even those who may not be outspoken on a day to day basis). We use NLP to recommend posts users and politicans based on previous likes and comments. To ensure quality, we have politician and proffessionals' comments at the top. Politicians can use this app to see trends and advocates for minorities.

How we built it

We built this platform using reactJS which communicates with a Django backend. Our database is made with PostGreSQL and for our post/article recommendation system we used a Gensim open-source library to build a LDA model.

Challenges we ran into

We’ve never tried Natural Language Processing, so learning what it does and how to use it was a little confusing in the beginning, but we quickly got over that hurdle.

Accomplishments that we're proud of

We’re proud that we were able to make our project within the allotted time frame. We had to learn a lot of new things like Django and NLP but we got it working and we hope that future iterations of it can make a change!

What we learned

One interesting thing that we learned was how to create a machine learning model to classify keywords into different topics. None of us have ever worked with natural language, so it was really cool when we got it to work.

What's next for ThinkOpenly

There are some more features that we look forward to add. One element we want to add is creating a calendar system to show when nearby political events and demonstrations will happen. We also want to add a system to verify politicians. These are just a couple of the improvements we look forward to making.