Workshop January 2005/events/workshopjanuary2005/Recent Changes to Workshop January 2005SmallWiki 1.0Software Composition Group, University of Berne, 2003Workshop January 2005/events/workshopjanuary2005/Wed, 07 Sep 2005 13:42:42 GMTSeptember 7, 2005 13:42:42.439 by admin from 213.219.168.176Workshop January 2005

Workshop

Wednesday, January 19 - ULB, Brussels

This workshop was jointly organized by the FNRS-FRFC funded Research Center on Structural Software Improvement and the IWT-GBO funded ARRIBA project. The goal of the workshop was to identify short and longer term opportunities for collaboration on research topics directly or indirectly related to these two projects.

Contents

Structure

  1. Presentations

    During the morning session 10 PhD researchers gave short presentations (10 minutes) on their short term future research plans.

  2. Identification of potential areas of collaboration Based on keywords identified during the presentations, we identified potential areas of collaboration, around which could be discussed in the afternoon session.

  3. Focussed discussion around specific research domains In the afternoon session, the participants were divided over different groups to discuss around three identified research areas: evolution, comprehension and engineering.

  4. Conclusion

Presentations

During the morning session 10 PhD researchers gave short presentations (10 minutes) on their short term future research plans. The goal of these presentations was not to explain their research in detail, but rather to have a general idea of the topics of interest of the different researchers, so that we could better identify potential collaboration opportunities with other researchers. Based on these presentations, for every person a number of keywords were identified. Everyone was also asked to write keywords on a post-it note and put them on a blackboard. These would be the input to a manual "cluster analysis" to identify the most active and overlapping research areas. Below you can find the list of presentations, some of which are downloadable, annotated with some keywords that were associated to these presentations by the moderator (Kim Mens).

Identification of potential areas of collaboration

Post-Its

Based on the post-it notes with keywords that were put on the blackboard during the morning session by all participants, Kim Mens and Roel Wuyts tried to identify the key areas of interest and overlap. This would serve as input to the afternoon session. They came up with the following topics:

  • Aspects

  • (Co-)evolution

  • Models (design / architecture / domain knowledge / graphs / ... )

  • Software analysis and transformation (where software may be interpreted as either code / models / dynamic information / ...)

  • Software mining and comprehension

  • Declarative meta programming

These topics formed a fully interconnected graph, meaning that several researchers were working on / interested in different combinations of these topics.

Other topics that were mentioned by several researchers were :

  • visualisation

  • legacy systems

  • case studies

After the lunch break, these areas were presented to the participants, with the goal of finding three or four topics around which could be discussed in group sessions in the afternoon. After some discussion, we decided to group the participants around the following problem areas:

  1. Evolution

  2. Comprehension

  3. Engineering

Focussed discussion around specific research domains

In the afternoon session, moderated by Serge Demeyer, the participants were partitioned over the three chosen problem areas (evolution, comprehension and engineering). The overall goal of the workshop was to identify potential opportunities for collaboration. To focus the discussion however, each group got the explicit assignment to come up with an attractive PhD research proposal for a (fictive) student starting his PhD research, including a concrete research plan for the first year.

The proposals they came up with were :

  1. Comprehension for aspect identification using language-specific knowledge

  2. View-based evolution

  3. Modularisation of model composition

    • ... is crosscutting

    • ... and requires aspects

In a second discussion round some participants were asked to switch groups and play the role of the student solliciting for the proposed PhD research topic. This allowed the groups to further refine their proposals.

Conclusion

It became clear that among the participants of this workshop quite some critical mass existed around the identified problem domains. Furthermore, the quality of the research conducted by the participants in these areas is very high, even at an international level. Therefore it certainly looks useful to organise a follow-up workshop where we could work out concrete collaboration topics. In fact we were thinking of organizing this wokshop recurrently with a frequency of 6 months.

This workshop was jointly organised by the

]]>