This sample implements a weather reporting system that can be used by the end users to mine temperature reports of some US cities.

(i)The application tracks the temperature of important cities in the US, available through a public web service at http://services.xmethods.com/ve2/ViewListing.po;jsessionid=hnXezdkbZATXgSu-vQtcxPuM(QhxieSRM)?serviceid=8.

(ii) A table function that takes a range of input values (zipcodes) as arguments is used to provide the resulting data of external web service invocation, as a virtual relational table.

(iii) Mining Current temperature: the end user can request the current temperatures of selected cities -- which will result in directly invoking the Temperature Web service  with the zipcodes of the selected cities -- and apply the full power of SQL including aggregate functions such as avg, min and max on the resulting data.

(iv) Mining Recently Collected Temperature: a database table is used to store temperatures collected since a determined period of time. A database batch job is scheduled to run every hour (or a determined period of time) and calls-out the Temperature Web service with the zipcodes of all the cities and store the data in a real database table. The end user can "mine" the recently collected temperatures of selected cities, from this table, using the full power of SQL including aggregate functions such as avg, min and max.

This tracking and mining approach can be applied to any "business data" available as external web services such as stock prices, scientific data, IRS tax tables, and so on.

For complete information on the architecture, downloading and configuring the demo, refer to the Readme.