DATAMIND

Philips Healthcare has a wide range of data available for its employees. This data however is hardly used, because it forms an unstructured overload to the employees. In addition, many departments are not aware of the data that is available already, leading to double work and missed opportunities.

A screenshot of the Datamind prototype

During the course of a two-week module, we have worked on a solution to stimulate knowledge sharing within Philips Healthcare. The result is DATAMIND: a smart suggestion system. DATAMIND improves the work-flow by getting the right information to the right person at the right time. The system also stimulates the creation of a social network based on knowledge sharing. A fully functional prototype of DATAMIND was made, using a fake windows environment. Below the full petrinet (information flow) within the Datamind prototype can be seen.

The full petrinet (information flow) within the Datamind prototype



About the System

DATAMIND uses a large database structure for storage, based on Philip’s servers. Relavant information is found through a contextually aware search algorithm, which analyzes the actions of an employee. By use of text-mining scripts, the system can analyze text in real time for interesting keywords. Search results from the database are filtered based on personal user information, and the relevant documents are shown on the user’s desktop. Active search, history views and changing the filter can be done in the main program of DATAMIND. Here dynamic visualizations make the user more aware of the available information and connections to other employees.







Background

DATAMIND is part of the CRISP project ‘Datafusion’. It was developed together with Fiona Jongejans, Koen van Ham, Josje Wijnen, Bert Bogaerts, Dennis de Klein, Martijn Peeters, Gabriele Tempesta and Bianca Wu.



A screenshot of the Datamind prototype