UAB Projects Team

Making great projects out of our University assignments



About us


We are a couple of computer science engineering undergraduate students at dEIC inside ETSE Engineering College, inside UAB campus in a town called Bellaterra, near Barcelona, Spain

Since we began to do all University practical assignments together as a team, we've put all our efforts in our projects so they become the best projects we can build before the deadline arrives. Usually those projects are timed to be developed between 2-4 weeks

We've created this website to offer the world our best projects and to keep our followers updated about new interesting projects

Unfortunately this is the last course in our enginering, so just few projects are about to come. You can check what we've done until now (note that most projects are private and pending to become public)


Bitcoin Payment Channels

We’re currently developing an amazing project as our final career project: a software that allows offchain Bitcoin Payment Channels so we can increase Bitcoin transaction throughput and improve Bitcoin scalability.

The idea is to allow nodes running our software create payment channels between them so that they can pay to each other with Bitcoin-valid transactions that aren’t sent to the blockchain (despite they are valid) until the channel is closed. This way, just the channel opening and close transactions would appear in the blockchain but thousands of transactions can happen between the parties meanwhile.

To implement the idea, we have to deal with game theory, Python 3.6 and low-level understanding of the Bitcoin transaction.

You can checkout how the repo grows and the project goes here:
https://github.com/uab-projects/btc-payment-channels

A Pythonian map-reduce implementation

We knew what big data was since last year we were taught in Distributed Systems how systems can be prepared to deal with lots of data and be able to deal efficiently with the data (and installed Hadoop!). That was the sysadmins point of view.

This year, in the Software Techniques and Architechtures subject, we’ve been taught what is the trendy term big data again, but now focusing on how to design a software architecture in order to deal with big amounts of data. Here is when map-reduce paradigm comes.

In this project, we’re asked to use whatever language we desire (we choose Python, of course*) and implement a map-reduce architecture that, as an example, counts the words or letters in a text file (a really big text file!)

Checkout how our projected ended looking: https://github.com/uab-projects/map-reduce

*it’s academical and we’re not worrying about high-end performance, otherwise we’d have chosen C++ or something compiled. (We know you were thinking that)

Sentiment Analysis with Twitter data

Another artificial-intelligence-cool-project from the KRU: Knowledge, Reasoning and Uncertainty subject.

Given a dataset with a few tweets (a million or more) as a ground-truth that we hope some hard-working human has been spent time on qualifying Tweets as positive or negative, allows to classify new incoming Tweets using bayesian inference.

You can check your account to see if you’re being positive on Twitter or not. Surprisingly most accounts on Twitter had more than 50% positive tweets (yes, even Donald Trump)

Checkout how we made it at:
https://github.com/uab-projects/bayesian-tweets