Mind Crawler

A PWA that turns your written thoughts into mind maps.

Comment

The Team

Tian Breznik - Front-end Development [Text Editor & Website] - https://www.linkedin.com/in/tianbreznik/
Mati Downes - Back-end Development [LTD word recognition] - https://www.linkedin.com/in/matti-downes-099689150/
Antonia Holslen - Back-end Development [API & Word2Vec] - https://github.com/Tonidor
Daniel Amador - Back-end Development [API & Word2Vec] - https://github.com/DanAmador
Brin Žvan - Design & Front-end Development [Brand guidelines & Website] - https://nirb.artstation.com/

Inspiration

Sometimes it's really hard to get to the bottom and boil your ideas down to the core. It is however, much easier to quickly jot down a thing or two.

Structures such as Word2Vec and LTA can analyze the relation of words so why not use that?

What it does

The application takes a text entry submitted by the user and runs it trough an analysis process which in turns clusters the words and themes extracted from the submitted text.

It uses the gained data to create a beautiful mind-map that allows us to collect and reflect on our thoughts more efficiently.

How we built it

The backend utilizes a database to store the entries and in turns runs them through trained Word2Vec model to predict establish the clusters and topics to which they form part of.

However no database starts full so it is on the front end side where the user submits their text. We built a website using Bulma to create a resource efficient site. The text editor and it's connection to the API was programmed in JavaScript alongside with the rest of the side there was a good portion of HTML 5 thrown into the mix.

Challenges we ran into

The first challenge we ran into was the lack of front end focused team members. None of us had much experience with is so we just went ahead and learnt as we went.

The second largest challenge was creating a system that combined Word2Vec with LTD which in turn should, in theory, give us a system that creates topics and categorizes.

Time.

The closer we got to the end of the hackaton the faster we realized that time and managed of that was our greatest enemy.

What we learned

From the most practical 16 hour javascript/html5 crash course to the fact that a good plan is essential to sucsess we mostly realized learned that 24 hour ideas should be byte-sized.

Built With

Share this project:

Updates