Machine Learning is a set of methods and techniques for constructing software systems automatically by analyzing only examples of the desired behaviour.
You do not write a program. You only provide examples of what you want. The program will be synthesized automatically.

In our lab we focus on engineering applications: have a look at our publications.

The Lab is directed by prof. Alberto Bartoli, assisted by prof. Eric Medvet, and hosts PostDocs, PhD students and LM, LT students: check the opportunities.
The Lab is part of DIA, University of Trieste, located in Trieste, Italy.

Follow us on Google+ Follow us on Twitter
Recent news Feed icon

We are finalist (again) at the "Human-Competitive" Awards!

24-06-2016Our work on automatic generation of regular expressions from examples (see this, this, this and this) has been selected for inclusion among the 8 finalists at the annual "HUMIES Awards for Human-Competitive Results produced by Genetic and Evolutionary...More

Our recent papers...

23-06-2016Very good news for the lab: three papers just accepted at prestigious conferences! These papers focus on very different application domains and propose approaches based on very different machine learning paradigms. Syntactical Similarity Learning by means...More

...and then IEEE Intelligent Systems !

23-03-2016Our last post was about acceptance on IEEE TKDE of our work on automatic regex construction and ended as follows: We are crossing our fingers as we hope to make another announcement soon... And now, yes, we have just received another acceptance notification from...More

IEEE Transactions (TKDE) paper!

21-03-2016Our "big paper" on automatic generation of regular expressions from examples will appear soon on IEEE Transactions on Knowledge and Data Engineering! We are really very proud of this result. According to the "State of the Journal Editorial" published...More

Active Learning of Regular Expressions

01-12-2015We have been working for several months on a significant improvement of our regex generator tool (source code): the ability to work in an active learning fashion. Rather than requiring the user to identify a set of examples describing the desired extraction...More