Integration of Rapidminer with Business Intelligence and Dashboard Applications
RapidMiner is a product used by quarter million users for their data science needs. In most datascience projects the final mile is delivering the results to a broader audience, the audience which can act on based on predictions. In RapidMiner we call this operationalization of Predictive analytics. As part of operationalization a very common practice is delivering the results via Dashboards, Reports or Visualizations. The following article describes some of the ways The following article describes some of the ways RapidMiner can provide highly effective solutions for these needs RapidMiner can provide highly effective solutions for these needs
1) Rapidminer Web Apps
2) Integration with Tableau
Rapidminer provides following three ways to integrate with Tableau.
a) Tableau Data Extract One way is use the "Tableau Data Writer" operator, this method allows creating Tableau Data extract file using RapidMiner workflows. Such processes can then be scheduled to push updates to Tableau data files thereby updating the Dashboards. Details about how to use this method are available here
b) Odata Service : Tableau provides ability to read data from an OData webservice. RapidMiner server provides ability to expose RapidMiner workflows as a webservice that can provide data using OData format. Details about creating webservices are available here and here. Ensure that when you create webservice chase the output format as Odata.
c) Tableau Web Data Connector : Tableau provides extensibility with with Web Data Connector. If this data source is preferred you can still use the webservices exposed from RapidMiner , by handling the data with web data connector API.Details about creating webservices are available here and here
3) Integration with Qlikview & QlikSense
RapidMiner provides atleast two ways to integrate with Qlikview.
a) QVX (Qlikview Extract) - This methods allows writing data out to Qlikview extract formats that can be used in both Qlikview & Qliksense applications.Such process can then be scheduled to push updates to Qlik Extract files, thereby uploading the dashboards.
Details are available here http://docs.rapidminer.com/server/advanced-topics/qlikview-integration/
b) Web Service - RapidMiner provides ability to expose any processes as a webservice. Each webservices can return data in various formats including XML, JSON, HTML table etc. Using this technique we can pull data directly from RapidMiner server based process into a Qlikview or QlikSense dashboard. Details about this method are available here
4) Other Applications
If webservices are not possible to be invoked from your BI application, you can still rely on RapidMiner to write its output to databases or files as needed. Your BI Applications can then get the data from these intermediate stores as needed.