Surety is performing system maintenance this weekend. Electronic date stamps on new Prior Art Database disclosures may be delayed.
Browse Prior Art Database

Dynamic Model Synchronization on DSMS

IP.com Disclosure Number: IPCOM000237358D
Publication Date: 2014-Jun-16
Document File: 7 page(s) / 68K

Publishing Venue

The IP.com Prior Art Database


This article describes a method to enable DSMS application to keep scoring incoming data without shutdown or suspending, even when the data mining tool DSMS leverages changes its data model. The method saves time of stopping and restarting DSMS, and ensures the incoming data is not lost during data model changing.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 7

Dynamic Model Synchronization on DSMS

Problem: When using DSMS to leverage data mining tools to do data scoring, if the data model changes, the DSMS application needs to be stopped and then restarted, so that data model file could be replaced when the application is stopped, and the new version of the file would be applied to the data mining tool and the DSMS after the application is restarted.

Known solution: none.

Why is an additional solution required: due to this problem,

1) it requires manual operation to stop the DSMS application, replace the data model file and restart the application, which costs extra time while processing data stream;

2) it may leads to data lose, if there is incoming data in input stream while the application is stopped.

Core idea:

1) enable the DSMS application to retrieve new data model file and refresh the data model;
2) enable the DSMS application to continue scoring data until the new model is applied.


1) it saves the time of the manual operation;

2) the incoming data stream could barely lose data.

In this invention, a mandatary operator named Data process operator with 2 threads provides the functions of the data scoring and model refresh for the DSMS application. Normally, this operator does data scoring when there is input from input stream for data scoring, and produces scoring result into the output stream. Another Notification operator (optional) could be configured in the DSMS application to feed an optional input stream of the 1st operator. The 2nd operator detects data model change, and notifies the 1st operator when the model changes. Once the 1st operator get notified, it prepares the new model, and meanwhile keeps scoring data. Once the new model is...