In our demo, we consider the data generated by a medical application as follows. Suppose a medical center that monitors key biological parameters of remote patients at their homes, e.g. using sensors in their bodies. The sensors periodically send to the center the patients’ health data, e.g. blood pressure, hydration levels, thermal signals, etc. For high availability, there are two or more sensors for each biological parameter. However, the data sent by sensors may be uncertain, and the sensors that monitor the same parameter may send inconsistent values. By using yet proposed approaches, a confidence value for the data sent by each sensor is estimated, e.g. based on their precision. According to the data sent by the sensors, the medical center analyses the health status of the patients. It also decides on the number of required equipments and human resources, e.g. nurses and physicians. For taking the decisions, the medical center poses its queries to a probabilistic database that maintains the probabilistic data, e.g. in our case ProbDB. An example of the issued queries is “return sum of the number of required nurses”. In our demo, we use the relations designed for the above medical application, and show the way by which the user can use ProbDB. Particularly, we demonstrate the following scenarios:

Data manipulation

In this scenario, we show how the user can insert the probabilistic data to the database by using a specific graphical user interface that supports the attribute-level model. We show how the data are maintained in the DBMS. We also demonstrate the way the user can update the inserted data.

Query processing

In this scenario, we show how the user can issue his/her queries. We show the different expressions that are provided by ProbDB for issuing probabilistic queries. The user issues some queries over the probabilistic health database, and ProbDB returns the results in a graphical user interface. The user can also choose some predefined queries and see the results.