Software Engineering for Smart Data Analytics & Smart Data Analytics for Software Engineering

User Tools

Site Tools

Lab "Model-Driven Software Engineering"

Dr. Günter Kniesel, Dr. Hamed Shariat Yazdi

General Information

The lab are part of the ICM track and intelligent systems track of the M.Sc. curriculum. It is aimed at students who want to specialize in software engineering or intelligent systems.

The lab is partly a follow up course for the MDSE seminar in the summer semester, which has layed the theoretical foundations for the technologies that will be used to define a new domain specific language for machine learning / deep learning. Participants of the advanced logic programming course who are familiar with the idea of data and meta-data will also easily understand the concepts of models and meta-models that are at the core of MDSE. Courses in the machine learning / deep learning domain are also good starting points, since they help understand the target domain that we will model.

Students who attended any such courses will be given priority in case that there are more lab candidates than available places. Other participants are also welcome but might need to invest more time into catching up with some basics your colleagues already have.

Model Driven Software Engineering

Increasing complexity of modern software calls for more effective and predictable approaches to software development. This need has led to the rise of a new paradigm: Modell-driven Software Engineering (MDSE). MDSE is being successfully used in many domains and is continuously evolving. There is a wide range of tools and technologies that use or support MDSE. Knowledge of MDSE concepts and tools has become an elementary skill for a software engineer.

Lab Topic

The lab will use MDSE technology (in particular Xtext and EMF) for generating a textual DSL for deep learning. The first aim of this language is to capture high-level concepts of deep learning in a uniform way. The second aim is to automatically map these concepts to various existing deep learning frameworks, thus relieving programmers from this highly coplex, error-prone and tedious task. The mapping will be ideally performed completely by the compiler of our new DSL.

The practical experience and the conceptual / technological insights gained during the lab are expected to help identify further options for improvement that could be pursued as master theses. Successful lab participants will be the primary candidates for these theses.

Place and Time

The lab will take place in the Römerstr. 164 building, room A125, each Friday from 09:00 to 17:00. Lab dates:

Meeting Date Advisor(s)
1 13.10 GK, HSY
2 20.10 GK, HSY
3 27.10 GK, HSY
4 03.11 –*– HSY ← GK available via Skype
5 10.11 GK HSY ← Start already at 8:30!
6 17.11 —— HSY
7 24.11 –*– HSY ← GK available via Skype
8 01.12 GK, HSY
9 08.12 GK, HSY
10 15.12 GK, HSY
11 15.12 GK, HSY
22.12 – free –
29.12 – free –
05.01 – free –
12 12.01 GK, HSY
13 19.01 –*– HSY ← GK available via Skype
14 26.01 GK, HSY
15 02.02 GK, HSY ← Final presentation & retrospective

The 2.2.2018 ist the last lab day. Adhering to the agile principles of “reliable delivery” it will really conclude the lab. You will present your final results and we will look back at what we achieved, what each of us learned, what we take with us as good practices and experience and what we take with us also as lessons learned about which things to do better and how.


For the time being we gather all information in our two Git repositories (docs, code).

Mailing List

  • mdse course lists iai uni bonn defill spaces with- @ . . - .”) ← will be set up after registration of participants

Teaching Staff

Who E-mail Tel Office
Dr. Günter Kniesel gk -at- cs uni-bonn de (0228) 73-4511 A107
Dr. Hamed Shariat Yazdi shariat -at- cs uni-bonn de (0228) 73-4506 A108
teaching/labs/mdse/2017/start.txt · Last modified: 2018/05/09 01:59 (external edit)

SEWiki, © 2024