-
Creating this app took resilience and hard-work from the whole team.
-
The software intelligently navigated social media to find reliable data sources to scrape data from.
Our mission: To connect communities globally in times of need.
MAS is a program made to connect people around the world by using social media to map man-made and natural disasters around the world in real time. Facebook Graph API was used to scrape Facebook posts. We also used the Twitter API to get statuses streaming with disaster filters. We used our homemade ML model using Tensorflow(trained on Google Cloud) to classify the statuses as disaster-related or not. Then we used the Microsoft Azure API to get location and name entities from the tweets. Using machine learning algorithms, we narrowed down the most accurate posts and classifies them based on the disaster (labels include earthquakes, explosions etc.). These events are now displayed in real-time on an interactive map.
Here, the user can select events to learn more about; the app redirects the user where they can click on social media posts/news articles pertaining to the disaster. The final feature implemented on this page is a place where the user can safely donate some money to those affected by such disasters and their information is secure thanks to use of Blockmason's Link API. In fact, we are THE FIRST to fully implement Link API in an app.
Businesses can gain lots from this service as they can subscribe to track different locations they want to follow. This is extremely beneficial as it can help them with supply chain management. Knowing what is happening in real time can help assess risk and loss which can aid businesses in proceeding in appropriate contingency plans if needed.
Using plenty of platforms to perform a vast variety of tasks proved timely to integrate with each other as the app has to be updated in real-time. Luckily, our teammates have diverse skill sets which allowed to harmoniously integrate our programs. Facebook permissions was a major road-block as the company had to approve of the app before scrapping can be completed safely which took 2-3 business days. We took on this challenge by only scrapping Facebook sites which the site allowed us to and this program was backed by some tinkering with selenium.
Coming into this hackathon, we wanted to explore new technologies while creating a product which can benefit and connect people from around the world. We structured our service around our mission and in the process, we maximized our team members' strengths and ultimately enjoyed this experience. We are all honored to be given the chance to do something great and hopefully this able to connect as many communities as possible.
Log in or sign up for Devpost to join the conversation.