[CogSci] Seeking for case partners: A new explorative human-data dialog technique

[Mauri Kaipainen]

Perspicamus is seeking for partners with needs to explore unstructured data to a project that aims to turn an innovative human-data dialogue technique (patent pending) into a product. 

The technique allows even non-data scientists to discover patterns of similarity through a sequence of intuitive and transparent human-data dialogs. At each step, the system provides the human agent with heuristics that assist in identifying the variable that has 1) optimal statistical characteristics while best 2) serving the analytic purpose. The chosen variable is applied as the criterion to subdivide the subdataset in focus iteratively into finer subclusters until an actionable level is reached. The analysis results in hierarchical cluster structures determined by explicit criteria and guaranteed relevance for the analytic intention.

Slide set: 

	https://www.dropbox.com/s/ui9qx5x1pgikkmd/IEEE%20Big%20data%202017%20Boston2.pdf?dl=0
	
Poster paper at IEEE BigData 2017, Boston: 

	https://www.dropbox.com/s/b69djhbudn048a2/Kaipainen%26Pitk%C3%A4nen%202017.pdf?dl=0



Mauri Kaipainen

CEO, PhD, senior scientist
Perspicamus Analytics
www.perspicamus.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cognitivesciencesociety.org/pipermail/announcements-cognitivesciencesociety.org/attachments/20180116/94ef2072/attachment.html>