Research   
   

15th of April 2011: So far, I've been reading a selection of papers trying to understand what MetaDepth is all about. After reading Deep Meta-Modelling with MetaDepth, I learned about the problem with strict meta-modelling, in which an element of a meta-layer is the instance of exactly one element at the upper meta-level. This style of meta-modelling has some limitations, in particular concerning the existence of only one kind of instantiation relation and the constraint of considering only two adjacent meta-levels at the same time. Hence, engineers are often forced to squeeze concepts into two meta-modelling layers that would naturally span several layers. This results in more complex and cluttered models. This problem is depicted in another paper, Reducing accidental complexity in domain models. Papers that offer solutions to this problem: Mapping between Levels in the Metamodel Architecture and Rearchitecting the UML infrastructure. In what follows is a list of papers I've read and will be reading in the following week:

  • Deep Meta-Modelling with MetaDepth (read)
  • Mapping between Levels in the MetaModel Architecture (read some parts)
  • Rearchitectuing the UML Infrastructure (to be read)
  • Reducing accidental complexity in domain models (to be read)

As I make progress on MetaDepth and strict/extensible meta-modelling, I will post updates on these matters. Another part of my reading assignment handles about Android. I'm still looking into several sources for more information on this subject. So, I guess I'll know more about Android in the coming week.

26th of April 2011: Been looking for some useful Android sources but couldn't find any scientific sources. Other sources that I've read and may be useful in the future:

The Busy Coder's Guide is an extensive introduction to several aspects of Android Development. It starts from the beginning, explaining what components an Android application is made of, how a project is structured, how layout works in Android etc. The other sources are hands-on guides to Android Development. They will come in handy when development for Android has started. However, in the near future I'll merely focus on (deep) meta-modeling and collaborative environments.

8th of May 2011: Read all papers concerning meta-modeling and started reading some papers prof. Vangheluwe gave me on collaboration. So far, I've read Co-ordinating Distributed ViewPoints: the anatomy of a consistency check. I guess I'll have to read it again in a few days, because not everything in the paper was crystal clear. I also haven't been able to relate the contents and subject of this paper to the overall view of my research (internship). I'll have to take a step back and think about it (or ask prof. Vangheluwe about it).

9th of May 2011: To have a better understanding of what needs to be done, I've created a small overview of my research internship and how my scientific article will look like. Every section has a relation to the articles I've read this semester. Those (sub)sections are probably subject to change.

  1. Introduction

    This section explains the contents of the paper and what will be discussed (problems with MM, deep meta modeling, metadepth).

  2. Traditional metamodeling approaches

    Introduction to metamodeling and the UML infrastructure

    • Rearchitecting the UML Infrastructure - T. Kuhne & C. Atkinson
  3. Problems with current metamodeling approaches/UML specification
    • Rearchitecting the UML Infrastructure - T. Kuhne & C. Atkinson
    • Reducing accidental complexity in domain models - T. Kuhne & C. Atkinson
  4. Deep metamodeling (a solution)
    • ProductType/Product problem example?
    • MetaDepth, DeepJava --> Potency
    • Can programming be liberated from the two-level style? - T. Kuhne & D. Schreiber
    • Deep Meta-Modeling with MetaDepth - J. De Lara & E. Guerra
    • Generic Meta-modelling with Concepts, Templates and Mixin Layers - J. De Lara & E. Guerra
  5. Conclusion
  6. References

31th of May: It's been a while since I've posted an update on my research domain. However, in the meantime I've been working on my article on deep meta-modeling. My current goal is to finish a draft of it by the end of this week. I'll have to decide with prof. Vangheluwe when we set the deadline for it exactly. I've put a main focus on the problems with deep meta-modeling and am currently reading about DeepJava and Metadepth (a paper I have received from prof. Juan de Lara, Generic Meta-modelling with Concepts, Templates and Mixin Layers). I particularly like the concept of potency, and am actually looking forward to using it in a real project. I hope to get it working in a basic Android prototype as soon as possible. The following picture describes the concept of potency:

This figure is extracted from Reducing accidental complexity in domain models. In short, it allows us to use deep instantiation. Each level, the potency value next to a clabject or attribute gets decreased by one, until it is equal to zero. At that particular level, we are able to instantiate and use it. For instance, the attribute picQlty is defined in the 'model metatypes' level and instantiated as a slot on the model level. If this sounds interesting: you can read more about it in my research article (will be linked in a post when ready).

Maintained by Philip De Smedt. Last Modified: 2011/05/31 22:57:11.