PhD Project Proposals by Andrzej Wąsowski

All PhD scholarships at IT University are granted in open competition. There is typically two main calls a year, one in Spring and one in Fall. Watch this space for announcements. Occasionally, it may happen that there is a call outside this schedule, but it would always be posted on the same web page.

If you consider applying for a PhD project under my supervision, please contact me first, before filing the application. Students in Copenhagen area are most welcome to ask for a personal meeting.

Automatic and Interactive Analysis Techniques for Concept Modeling

The project is aiming at exploring a number of open problems in analysis for conceptual models; here anything between simple feature trees to class diagrams. The aim is to design techniques that efficiently aid software engineers in creating, manipulating, diagnosing, measuring and/or instantiating these models. The techniques will presumable be based on wide range of technologies available in AI, theorem proving, model-checking and operation research community.

Contact Andrzej Wąsowski for more details.

Smart Modeling Environments for Universal and Domain Specific Modeling Languages

We observe a growing interest in introducing intelligent user support in modeling environments, with particular emphasis on domain specific modeling environments. Authors in various traditions (graph transformations, logic programming, constraints and optimizations, Bayesian reasoning) are trying to bring AI support in modeling environments to a new level.

In this context there is several areas in which a PhD project could well contribute, for example:

1. Building a semantics preserving model editor for modeling languages defined using a metamodel in Ecore plus a subset of OCL (or using a similar technology, like MetaEdit+ setup, or Microsoft DSL Tools). It would be desirable to understand properties of guidance of such an editor, striving for exhaustive, and valid guidance. (See Kuzina et. al in MODELS'08)
2. Address the problem of inconsistency containment in such a modeling editor, so that meaningful, and possible high quality guidance can be provided even for models that contain inconsistency.
3. Address the problem of consistency recovery, by helping the user to reach a consistent state in the modeling process.
4. Incorporate probabilistic inference methods into guidance mechanisms of modeling environments.
5. Explore possibilities of supporting meta-model editing in languages like Ecore, in a completeness presering fashion. (See Kuzina et. al in MODELS'08)

Contact me for references to relevant literature. The list of topics is by no means exhaustive. Neither it is required that all of them are addressed in a single PhD project.

An ideal candidate has strong background in programming languages and model driven development, with a moderate background in AI (logics, constraints, Bayesian inference).

CONTACT: Andrzej Wasowski

From Model-driven Development and The Software Product Lines Paradigm to Automatic Support for Programming with Frameworks

New programming languages, technologies, standards and libraries appear, become popular, and die quietly almost every week. Given this explosion on the market of ideas, it has become very usual that programmers work with tools and technologies they barely know. These programmers need to spend a lot of their development time on learning, by reading manuals, and searching the web for examples and discussions.

This project aims at automating some part of that learning by extracting the variability/commonality information from applications using programming frameworks, and using it to guide the developer in her work.

The project includes mining the source code for patterns API use, program analysis, model-synthesis, interactive guidance, and automated synthesis of test cases. It is expected that the results of the work will be implemented and validated within the Eclipse project.

An ideal candidate has strong background in programming languages and model driven development, with a modarate background in AI.

CONTACT: Andrzej Wasowski