Our team was heavily inspired by artificial intelligence and machine learning, and using the programming skills available to us, we thought we would be capable of constructing a simple language-analysis model to analyze the sustainability of an action based on some core principles, namely the UN's sustainability "layer cake" model.
What it does
The application is comprised mainly of a simple in-and-out chatbot that asks for information like recyclability, economic efficiency, and other details that affect the sustainability of an action. The chatbot is then capable of ranking the sustainability as a weighted percentage, 0 being unsustainable and 100 being perfectly sustainable. Additionally, a clean graphical user interface allows for interaction with the user without having extensive knowledge of the command line and having to use a clunky command prompt for I/O.
How we built it
The GUI was constructed first; on top of that, an
intelligence module was created, which, upon initialization, is imported. The GUI was programmed to read and write to the AI, which received and analyzed the information from a user prompt.
Challenges we ran into
Artificial intelligence is always difficult, and while the language analysis came mostly clean, the state processing and sustainability analysis were a definite slowdown. However, in the end, a simple proof-of-concept was completed.
Accomplishments that we're proud of
The GUI is nice, clean, and code-readable, and the AI works and is able to detect objects of interest in a sentence. These are certainly achievements to be proud of.
What we learned
Backend and functionality should be programmed first; then UI and UX can be implemented. Solving the difficult problems first is more likely to return better results.
What's next for SustAInability
Post-proof-of-concept, a fully-functional prototype could be constructed that could fully analyze a statement and return a sustainability score.