Main products#

  • d3web-Core: The reasoning engine and persistence components for problem-solving knowledge including decision trees, (heuristic) rules, set-covering models and diagnostic flowcharts.
  • KnowWE (Knowledge Wiki Environment): As the successor of KnowME the system KnowWE (Knowledge Wiki Environment) offeres a web-based wiki front-end for the knowledge acquisition and supports the collaborative engineering of knowledge at different formalization levels. Strong problem- solving knowledge is mixed with corresponding text and multimedia in a natural manner. The knowledge base can be flexibly structured by distributing the particular knowledge modules over a collection of linked wiki articles, each covering a particular aspect of the domain.

Applications#

The applications listed below are applications using either d3web-Core, or both, d3web-Core and KnowWE.

Open Source Applications#

  • KnowWE-Headless-App: An version of KnowWE without the normally underlying web server. It can be used to produced d3web knowledge bases from wiki pages in a batch process.
  • ProKEt: Prototyping of knowledge-based dialog systems.
  • WUMPS: Knowledge acquisition based on Office documents.

Free Applications#

  • Mobile-App: Application allowing to use d3web problem solving knowledge in a stand alone application.
  • Droid3web-App: Application allowing to use d3web problem solving knowledge on your mobile device.

Commercial Applications#

  • KnowledgeBaseBook: Generates a DocBook XML file of an executable d3web knowledge base, i.e., the knowledge base book. The DocBook XML can be used for further transformations. The application produces a PDF printout.
  • Testing-App: Allows to continuously run specified tests on a d3web knowledge base in development.

Modules (under construction)#

Modules are the different parts of the main products and applications

Complex modules#

Licenses

  • add articles for the different licenses and link them
  • d3web-KnowWE-Releases contain all used licenses...
Albrecht Striffler
Describe the actual modules
  • Create new articles for complex modules with multiple plugins
  • Use ontology descriptions for "included plugins", "quality level", "license"
  • Describe/define quality level
  • Simple modules are not defined manually... all plugins that are not part of a complex module are simple modules aggregated on one page, probably using a sparql query or an own KnowWE-Plugin
Markus Friedrich

Module-Plugin implementation

  • Jenkins-Plugin to upload Plugin/Module information
    • https://wiki.jenkins-ci.org/display/JENKINS/Extend+Jenkins
    • SVN: https://svn.jenkins-ci.org/
    • Download jenkins war and add to tomcat
    • Get hello world plugin and import into new workspace
    • Plugin execution not covered by lifecycle configuration does not matter, it is just ugly
    • Maven install hello world plugin
    • Copy hello-world.jar to WEB-INF\lib
    • Copy hello-world.hpi to WEB-INF\plugins
    • Start tomcat/jenkins
    • Hello world plugin will be available as post step option
    • Start implementing new plugin to upload build data to d3web.de page as a attachment
    • Some additional code snipplets on how to upload files to a KnowWE Action
    • Provide information needed for module pages
Alex Legler
Module-Plugin implementation
  • Design ontologie about module information. We want for example the following information per module:
    • What maven modules / svn projects are part of this module (will be defined on the module page with turtle markup)
    • What projects are not in any complex/aggregated module
    • What are the last x changes to any project of a module
    • What is the heat of the module (changes/day in the last month?)
    • What are the test stats for the module
    • General stats, if possible? LoC, Classes, BuildNumber
    • Authors (people who created modules, changed them)
    • More interesting stuff?
  • Write KnowWE-Plugin to add data provided by Jenkins-Plugin to the rdf2go model.
  • If there is time: Write KnowWE-Plugin to render module information from rdf2go model in a nicer way. If there is no time: Use default SparqlRenderer
Sebastian Furth

Product Ontology#