System and a method for reasoning and running continuous queries over data streams

Patent metadata

Field	Value
Publication number	US-9990403-B2
Application number	US-201314387540-A
Country	US
Kind code	B2
Filing date	Apr 12, 2013
Priority date	Apr 12, 2012
Publication date	Jun 5, 2018
Grant date	Jun 5, 2018

Abstract

Disclosed is a method and a system for stream reasoning a plurality of data streams. The system comprises a processor and a memory coupled to the processor. The processor is capable of executing a plurality of modules embodied on the memory. The plurality of modules comprises an event module and a application managed window module. The event module is configured to receive a data stream associated with an event from a stream reasoner application. The data stream provides factual information about the event. Further, the data stream comprises a request. The request may be an add request or a delete request. The application managed window module is configured to insert the request associated with the event or delete a prior request associated with the event from the memory based upon a type of the request.

First claim

We claim:

1. A stream reasoner comprising:
a hardware processor; and
a memory coupled to the processor, wherein the processor is capable of executing instructions stored in the memory, the instructions comprising instructions to:
receive a data stream associated with an event from an intermediate application, being a state machine comprising a set of states, positioned between a user device and the stream reasoner, wherein the data stream provides a time-ordered set of facts about the event, and wherein the data stream comprises a request, and a type of the request is determined as one of an add request and a delete request based on a knowledge packet associated with the data stream received from the intermediate application; and
in response to determining that the request is the add request, insert the request associated with the event; and
in response to determining that the request is the delete request, delete a prior request associated with the event from the memory, wherein a state of a window, representing at least an order of all received requests, corresponding to the add request or the delete request is updated by an application management window module in the stream reasoner, wherein the instructions are further configured to:
receive a plurality of incremental queries, wherein each incremental query corresponds to a single event and assesses one single event at a time, wherein the plurality of incremental queries is supported by parameterized queries to run each incremental query multiple times with different parameterized values;
provide a logical reasoning about the event from the data stream and background knowledge using a set of rules, wherein the background knowledge supplements additional static information associated with the event;
execute the plurality of incremental queries on the logical reasoning, the data stream, and background knowledge, wherein a corresponding state of the intermediate application is updated incrementally with results of the executed plurality of incremental queries; and
combine the results of the executed plurality of incremental queries with the background knowledge to detect patterns associated with the event.

2. The stream reasoner of claim 1 , the instructions are further configured to:
configure additional sources to determine the background knowledge corresponding to the data stream.

3. The stream reasoner of claim 2 , wherein the additional sources are heterogeneous knowledge sources.

4. The stream reasoner of claim 1 , wherein an expiration of the request is dependent on at least one of time, count, and the type of request.

5. The stream reasoner of claim 1 , wherein the logical reasoning is provided based upon a reasoning technique comprising one or more of a logical reasoning, a deductive reasoning, a rule-based reasoning, an inductive reasoning, and an abductive reasoning.

6. The stream reasoner of claim 1 , the instructions are further configured to:
receive query fragments for the reasoning, the data stream, and the background knowledge;
execute the query fragments simultaneously on the reasoning, the data stream, and the background knowledge; and
determine a final result based upon the execution of the query fragments.

7. The stream reasoner of claim 6 , the instructions are further configured to:
execute the query fragments in one of a parallel mode and a sequential mode.

8. The stream reasoner of claim 1 , wherein the knowledge packet is organized as a set of triples, and wherein each triple comprises a subject, a predicate, and an object, wherein the inserting the request corresponds to inserting the set of triples in the request for addition to the memory, wherein the delete request corresponds to deleting the knowledge packet immediately from the memory or schedule the knowledge packet for deletion from the memory.

9. A method, implemented by at least one computing device, for controlling a request in a stream reasoner, comprising:
receiving a data stream associated with an event from an intermediate application, being a state machine comprising a set of states, positioned between a user device and the stream reasoner, wherein the data stream provides a time-ordered set of facts about the event, and wherein the data stream comprises a request, and a type of the request is determined as one of an add request and a delete request based on a knowledge packet associated with the data stream received from the intermediate application;
in response to determining that the request is the add request, inserting the request associated with the event; and
in response to determining that the request is the delete request, deleting a prior request associated with the event from the memory, wherein a state of a window, representing at least an order of all received requests, corresponding to the add request or the delete request is updated by an application management window module in the stream reasoner, wherein the method further comprising:
receiving a plurality of incremental queries, wherein each incremental query corresponds to a single event and assesses one single event at a time, wherein the plurality of incremental queries is supported by parameterized queries to run each incremental query multiple times with different parameterized values;
providing a logical reasoning about the event from the data stream and background knowledge using a set of rules, wherein the background knowledge supplements additional static information associated with the event;
executing the plurality of incremental queries on the logical reasoning, the data stream, and background knowledge, wherein a corresponding state of the intermediate application is updated incrementally with results of the executed plurality of incremental queries; and
combining the results of the executed plurality of incremental queries with the background knowledge to detect patterns associated with the event.

10. The method of claim 9 , further comprising:
configuring additional sources to determine the background knowledge corresponding to the data stream.

11. The method of claim 10 , wherein the additional sources are heterogeneous knowledge sources.

12. The method of claim 9 , wherein an expiration of the request is dependent on at least one of time, count, and the type of request.

13. The method of claim 9 , wherein the reasoning is provided based upon a reasoning technique comprising one or more of a logical reasoning, a deductive reasoning, a rule-based reasoning, an abductive reasoning, and an inductive reasoning.

14. The method of claim 9 , further comprising:
receiving query fragments for the reasoning, the data stream, and the background knowledge;
executing the query fragments simultaneously on the reasoning, the data stream, and the background knowledge; and
determining a final result based upon summation of the result of the execution of the query fragments.

15. The method of claim 14 , wherein the query fragments are executed in one of a parallel mode and a sequential mode.

16. The method of claim 9 , wherein the knowledge packet is organized as a set of triples, and wherein each triple comprises a subject, a predicate, and an object, wherein the inserting the request corresponds to inserting the set of triples in the request for addition to the memory, wherein the delete request corresponds to deleting the knowledge packet immediately from the memory or schedule the knowledge packet for deletion from the memory.

17. The method of claim 9 , further comprising processing the data stream using a combination of a reasoning t

Assignees

• Tata Consultancy Services Ltd

Inventors

• Mukherjee Debnath
• Misra Prateep
• Banerjee Snehasis

Classifications

• G06F16/24568 Primary

  Data stream processing; Continuous queries · CPC title

• G06N5/022

  Knowledge engineering; Knowledge acquisition · CPC title

• G06N5/04

  Inference or reasoning models · CPC title

• G06F17/30516 Primary

  Physics · mapped topic