Peak tweets tracking with time
First inspired by wanting to know what kinds of events might be taking place around me. Then, was able to find out who could be there, how many people might be there, where else are these users be relocating to with time, and potentially predict future events.
What it does
Workflow goes like this: first, search on Twitter or other SNS applications about certain key words you might be interested in looking into. It'll collect all the data about the users' real-time info into a txt file. Then what happens next is data processing to get rid of all the unwanted words, keep track of the key nouns and objects to create a new processed txt file; geo-coordinates of where the users' tweeted location is also made from which heat map is created on google map to show where users were. Many more future applications are left to be done. But this work shows great potential for keeping track of the certain people who may seem in danger, or who may be dangerous to someone or a group.
How I built it:
Mostly python programming, searching into several APIs.
Challenges I ran into
It was challenging to figure out how to process the collected tweets to make sure that what I collect are relevant data for my search. I have still yet to improve on my confidence level of my collected tweets.
Accomplishments that I'm proud of
This software could actually tell me when and where events I want to look for are happening at.
What I learned
How to use social media API servers to collect huge amounts of data, with which I could get out some pretty nice information about what's going on in the real-world and how are people reacting to those events in real time.
What's next for Social events tracking based on Social Media Records
Expand onto other Social Media App data tracking to potentially Instagram, Facebook, Youtube, dating apps in order to improve on the data collected.