Hi Pablo, The main goal of this task is developing an algorithm for test selection. If you succeed at it but don't have time to create interfaces, it should be very easy for us to fix it later. If, on the other hand, you will end up with the complete system but without the efficient enough algorithm, the whole thing will be useless. So, *please continue with the research and improvement of the algorithm*. Without going into detail of the new approach that you've taken (I'll do it later), if you are now looking for ways to improve the recall and to keep it high after a big number of simulation runs, I would recommend to analyze, maybe even one by one, the 'misses' -- failures that the current algorithm does not catch, and see why it happens in every case. It might give you some ideas where to direct your main effort: whether it's because the tests are not in the "dictionary" at all, and hence we need to find a way to fill it better and update regularly, or it's because the tests have too low metric and hence don't make the top of the queue, and then the metrics might need some improvement, or there are some other reasons which we don't even consider. Regards, Elena On 04.07.2014 11:28, Pablo Estrada wrote:

Hello Elena, I understand that the recall that we are achieving so far should still be improved; nonetheless, since we are already in the second half of the summer, I was wondering if we should spend some time designing the interfaces for the implementation of the project; such that I can start building them (even though the research phase can't be marked off as completed yet).

Regards Pablo