Seeing the increasing amount of “fake news” present in our modern world, I wanted to make a program that combats this problem.
What it does
NewsStream starts with a QR code, which leads the user to a text message screen to a certain number with the auto filled box: “Subscribe.” Once the user sends “Subscribe,” the program asks the user about his/her interests. The user will input their category of news interests, such as technology and politics. In addition, the program will scrape the country of the user to more accurately provide important information. Once these steps are completed, the program will send daily, verified, current news updates to the user at 7AM local time. The user has the following options: unsubscribe, add interests, remove interests, and view interests.
How we built it
I built the program using the following technologies: Heroku - This project is currently hosted on a single Hobby sized dyno, mLab - Sandbox tier MongoDB attached as an add-on to the Heroku, Google Dialogflow - Basic machine learning backed by Google allowing us to deliver a conversational experience to our users, and Twilio - SMS messaging API used in conjunction with Dialogflow. I coded these different systems using Python and Java.
Challenges we ran into
Training Google Dialogflow and Twilio to work in conjunction and create a conversational experience with the user was a troublesome task at first. In addition, due to the restrictions of Google Dialogflow, the text message service does not function correctly for specific phones. In the future, I would purchase a more adaptive and comprehensive system to use.
What's next for NewsStream
NewsStream has major room for global growth, as the product is adaptive to a large audience: everyone in the world with phones. In addition, the program is currently able to provide news to every country in the world. In the future, I would stop using Google Dialogflow as my main text messaging system to allow for this growth to occur.
Log in or sign up for Devpost to join the conversation.