Meetings   
   

12/15/2011 - We can take two avenues RDF and model transformations on RDF, or hierarchical graph matching using scope to speed up matching and reduce complexity.

  • Explore theoretical aspects of scoped graph rewriting
  • RDF - highlight difference in performance, raw iGraph vs operations on RDF store
  • Setup RDF store
  • Will work on implementing RDF support in tCore
  • Explore complexities that are possible.
  • Cluster access

01/11/2012 Expecting code from Eugene and Simon Vanmierlo. This will allow to do transformations in standalone mode.

  • Need to plug in RDF store into AtomPM or standalone
  • The fastest would be to do in memory RDF transformations
  • Later the ideal for scalability would be depending on the load to switch from iGraph to RDF in-database transformations using queries
  • Work with Levi to develop transformations for power window
Python to JavaScript complexity can be deprecated. Expecting report and presentation from Simon Vanmierlo about T-Core in commercial tool.

02/01/2012 Starting transformation work, State charts to Petri net.

  • Need to improve the profiler and instrument more places. Min, Max. Same entry but different exit must yield separate records.
  • Still trouble using Simon's example, will look later, profiling in Atom PM is working fine.
  • Logging server format change and retrieval for analysis purposes.

02/12/2012 RDF, Transformation benchmarking, Power window transformations.

  • By June 25th need to demonstrate RDF to IBM and analyse the performance.
  • Thorough benchmarking of the transformations, using relevant synthetic, generated models and transformations.
  • Concrete syntax manipulation. Union of models using named links?

02/29/2012 Commited for June 25th IBM meeting deliver RDF plus performance analysis of the existing system

  • PN analysis inside AtomPM as a project

04/11/2012 Paper for Bremen, RDF store transformations or modular PN analysis. Results by end of may, abstract June 12th, paper June 18th.

  • RDF store transformations, transactions, search plans, efficient querying.
  • Query only what you need. Compile rules into one query?

10/22/2012 Priorities:

  • Scoping specification in AtomPM rules such that it results in naive scoping application, Corsica fire example.
  • Pull out transformation out of AtomPM so that we can have huge examples for RDF case (RDF Himesis graph generation for Corsica example without using transformations that generate for example). Same applies to scoping.
  • Implement efficient scoping. Either simply make a light copy of a subgraph and apply VF2 or indexing hash table to provide additional heuristic for matching algorithm (can be your own).
  • Start writing thesis (introduction). Template. RDF will be scalability chapter within Efficient and Usable Transformations topic.
  • Work on extending transformation language such that we can have arbitrary transformation languages(ideal) or at least take care of pivots, match sets and iteration over them, etc.
  • Perhaps RDF paper needs to be changed with really huge examples of Corsica after we can take out transformations out of the AtomPM GUI.
  • SOCS doc. apply by January. NSERC September, need TA 522? | IWT - deadlines, requirements. Hans topics, funding.

11/06/2012 CASCON.

  • Batch RDF transformation script, rather than message exchange. See
  • How do we convert attribute constraint verification into SPARQL?





Maintained by Maris Jukss. Last Modified: 2011/12/20 06:03:21.