Skip to content

Dalvonic - CNN model semantic analysis + Tweets + Pretty Viz

License

Notifications You must be signed in to change notification settings

AHAAAAAAA/HackCU2016

Repository files navigation

Dalvonic

Semantic analysis + Tweets + Pretty Charts

#Description

With the explosion of the Internet, the dissemination of information and opinion has been democratized significantly. Opposed to a top-down information flow, there's now a bidirectional flow. Tools like Twitter, FB, and other services have served to empower activists and regular people to challenge the state and media narratives that are the primary opponents of free speech. Through our team's experiences with information wars and censorship in the Middle East, US, and Central America, we've identified the remaining challenges to freedom of speech to be three-fold: self-censorship, lack of decent Internet access, and echo chambers.

We chose to tackle the problem of echo chambers and, particularly, in-grouping. Despite the onslaught of massive amounts of information, services like Twitter, Reddit, Facebook, and other blogging services are highly personalized. Through social network graph analyses, it's very strongly evident that people follow and remain exclusively in the company of very like-minded individuals online. We aim to expose users to people who share opposing views to them and prompt users to consider information from ulterior, opposing sources based on sentimental and subjectivity analysis of their tweets (with a CNN classifier).

Quick setup:

Install Flask: pip install flask

Usage

python app.py

Dependencies

pip install: nltk, tweepy, whatever module is missing. Trust me.

About

Dalvonic - CNN model semantic analysis + Tweets + Pretty Viz

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published