The Big Picture

Inspiration

Our floor has a very lively group chat and we wanted to extract as much personal information as possible (and find embarrassing moments/ statistics).

What it does

It parses groups chat and then organizes it into clusters where each clusters is separated by 20 minutes of silence/ inactivity. Then it outputs statistics about the clusters such as most talkative, most lonely (last message of a cluster) as well as statistics about the chat as a whole.

How I built it

Used JavaScript to parse input messages and run analysis on it. The use html/css and 3d to display the data in graphical form.

Challenges I ran into

• Getting messages from Facebook and parsing it
• How and what data to give statistics for
• Using 3d to display visualization of data

Accomplishments that I'm proud of

• Cleanness of website
• Implication of 3d to visualize data

What I learned

How to implement 3d to visualize data

What's next for The Big Picture

• More graphs and analysis
• Search capabilities for custom statistics
• Make the bars on the graph proportional to the time interval of the cluster that it represents
• Speed up analysis of longer chats

Built With

• 3d
• css3
• html5
• javascript
• natural
• nlp-compromise
• sum