Some progress headlines:
- September 6th, 2010: First contact from José to Michael regarding the idea.
- September - October 2010: a couple of personal talks and many emails end up setting up and submitting the proposal.
- February 7th, 2011: the proposal is accepted for funding; some decisions about personnel, timing, and details about a couple of main events are taken.
- March to may 2011: initial discussions; first visit dates fixed, roundtrips and hotels booked...
- June 2011: Javier and José visit KNIME for several days (both U Konstanz and Zurich headquarters); José gives a presentation to the KNIME team and also an academic talk open to the whole department of U Konstanz, very well attended, as part of the training subprogram of STARK; Javier receives personalized training on the usage of KNIME internals.
- July 20th, 2011: His academic duties for the year finished (to the possible extent), Javier travels to KNIME to stay the whole summer and develop the main part of the project; one of the first tasks is the setup of this web page.
- September 1st, 2011: José is visiting back. The new main node has been useable for some days now, but results are not yet fully as we want them; we are chasing after small details of all sorts, and discussing the precise specification of related nodes that could simplify the usage as much as possible. One piece of bad news, though, is that the new implementation is still quite slower than we planned to have by now.
- September 8th, 2011: The new node works as intended, except for speed. Further fine tuning to be done. José and Javier are back at Cantabria, as there is no such need of interaction for the remaining tasks. Planning for the Closing Event of STARK, which will be a session devoted to this Harvest initiative within the Second Workshop on Applications of Pattern Analysis, october 19-21, which will feature presentations by Tobias Kötter, José Balcázar, and Javier de la Dehesa.
- The STARK session took place as intended. Yet, first, a bit of underspending left some money available, and, then, somebody around is never satisfied, so, further thinking was invested along the academic year 2011-12, and Pascal allowed for spending to take place along Pascal Year 5, if within budget. Diego, Cristina, and José are still working on the node. Iris Adä is helping us with the logo. Thorsten Meinl is helping us with the mechanism for adding the new node to any KNIME installation.
- A communication explaining the major issues behind yacaree, both in Python and in Java for KNIME, was presented at Int Conf on Formal Concept Analysis in Leuven in 2012.
- Major milestone: The first public version of the Yacaree Associator node is made public and is advertised at the KNIME Open Source Days in september 2012. But further versions will come, with an improved logo, with better logging, with functionality that the Python version already sports (implications) and the KNIME node still does not, and, hopefully, with a faster closure miner on which we have been working hard for quite some time now!
