Activity data level variable update fixed #36

fixed issues #36

Made faster, packages renamed and preparations for UI

• IMPORTANT: package structure has been changed in this version. If you used an older version make sure to update your code imports and maven/gradle dependencies.
• Speed update through using caching in Spark applications
• fixed a bug for activity level processing (sometimes variable data was missing in output)
• preparations for bpmn.ai-UI that will also be released on github

Docker support added

bpmn.ai can now directly be run via Docker.

Pipeline refactored

In this release the pipeline has been refactored to reduce the number of dependencies between the steps of the default pipelines to a minimum.
Configuration files are now created per Spark application (by adding a suffix to the filename), so that multiple configurations can exists in the same working directory next to each other.