wiki:Home

Welcome to the RECAA Project !

AutoModel The Idea

Improving
Requirements
Engineering
The Idea

Automatic
Model
Extraction
SaleMX
Blog

Publications
Tools
RESI
REFS
AutoAnnotator
Tools

Sale2UML
Sumox
Thematic Roles

Publications
Conferences Literature
&
Theses
NLRP
Benchmark
Contact
Team Members
Technologies
&
Concepts
Downloads
Links
Eclipse Update Site
Open Tickets
Internal
FAQ
Examples

The Research Field: Requirements Engineering Described

A typical requirements engineering process is depicted in the following figure.

https://svn.ipd.kit.edu/repos/koerner/mx/public/img/REProcess.png

The requirements analyst is responsible for coordinating these processes while improving and documenting the requirements. This approach is iterative and has the following steps

  • Discuss the requirements
  • Elicitate user needs, document requirements, verify requirements
  • Created (semi-formal) model of requirements
  • Hand-over model to software architect
  • Architect makes changes on the model
  • Changes have to be re-discussed with the stakeholder and changes need to be checked for compliance with the customer's wishes.

Requirements are often visualized using domain models which are used for the software development process. UML domain models are created from natural language requirements specifications. The analyst then uses domain models to verify and rectify the stakeholders' input which was gathered in discussions and interviews. These models are the blueprint for the software architect who might make changes to the models himself. These changes are usually not automatically fed back to the customer.

RECAA

After having completed some [tagged:theses theses] at our chair, the two projects AutoModel and SALE Model Extractor (SALE MX) found together founding the Requirements Engineering Complete Automation Approach (RECAA) team. We try to integrate our work into one easy-to-use product which can support requirements engineers throughout the entire RE process. In doing that, we try to automate as much of the work as possible, to enable the requirements engineer to focus on the tricky parts of RE.

AutoModel

AutoModel is an approach to improve the requirements engineering process during software development. It deals with natural language descriptions, their flaws and last but not least their conversion into UML. AutoModel is an add-on to already existing requirements engineering processes.

An overview of the complete AutoModel system and its subprojects is provided here.

SALE Model Extractor

The SALE Model Extractor (SALE MX) is a suite of natural language processing (NLP) tools. It has the ability to extract a UML domain model from an annotated text written in natural language to form a starting point for the Model Driven Architecture (MDA). The text must be annotated manually at the moment with a annotation language called SALE (SENSE Annotation Language for English). Even though SALE has been designed for English texts, it can be used for other languages aswell, e.g. German or Hungarian.

An overview of the complete SALE MX system and its subprojects is provided here.

Internal Documentation

If you're part of the team - or if you want to dive into our world - you should have a look at the internal documentation. We provide Common Eclipse, a fully featured distriution of our work. Try annotating a text using our editor plugins for Eclipse or improve a specification with RESI (using openCyc).

Visit the [tagged:internal "internal" pages] for additional information that might help during writing your thesis.

Last modified 5 years ago Last modified on Feb 6, 2013 8:47:33 AM