Inspiration
It's notoriously hard to get someone to conceptualize the same thing as you, when you say something such as "5 o'clock" in conversation, someone's experience with language and the context they brings in rather quickly determines their mental image. This problem gets exponentially more complicated when trying to communicate around knowledge work, research, and other domain specific topics such as trying to onboard a coworker to a new project, or figure out who to assign a task to, or know which process is most important to an organization's success.
What it does
Threshold visually represents ontologies from different data sources, in this case Pocket and GitHub. It does this in a way that assigns a physics system to your thoughts and the data you interact with daily. In this way, we can show how both an individual and an organization can benefit from seeing their vocabulary as a weighted graph, built off of data, rather than being limited to human explanation.
I hope that both the visual component and the underlying graph structure can allowing for exploration of the mental models of other individuals, allowing for communication and eventually agent based interactions that extend beyond what traditional communication mediums allow for.
How I built it
Flask, d3, js
Challenges I ran into
Hard to attract teammates this time around given the remote nature. Also d3 is pretty intensive so loading lots of data in a single view really slowed down development at various points
Accomplishments that I'm proud of
I was able to build a demo! The Pocket sourced graph in particular I feel like shows a representative interest graph. While the GitHub example show how a project split into two factions, a useful insight for someone like a project manager
What I learned
How to use d3.js. How to distill complexity of large data sets.
What's next for Threshold
Adding a user system, tying together the ontologies for multiple data sources, adding some teammates. Figuring out a monetization strategy...
Log in or sign up for Devpost to join the conversation.