Method for generating processing specifications for a stream of data items

• US 20110125762A1
• Filed: 12/04/2009
• Published: 05/26/2011
• Est. Priority Date: 11/24/2009
• Status: Active Grant

• Current Assignee: Software AG
• Original Assignee: Software AG
• Inventors: Hafermann, Udo, Schöning, Harald
• Granted Patent: US 9,286,353 B2
• Time in Patent Office: Days
• US Class Current: 707/749
• CPC Class Codes: G06F 16/24568 Data stream processing; Continuous queries

Abstract

The present invention concerns a method for generating at least one processing specification (30) for a stream (10) of data items (100a, 100b, . . . , 100n) captured by a sensor (5), wherein the method comprises the following steps:

• a. generating a plurality of proposed window specifications (20) and generating the at least one processing specification (30) based on at least one of the proposed window specifications (20);
• b. wherein the plurality of proposed window specifications (20) is generated based on a similarity metric adapted for identifying similar pairs of data items (100a, 100b, . . . , 100n) in the stream (10) of data items.

Claims

1. A method for generating at least one processing specification (30) for a stream (10) of data items (100a, 100b, . . . , 100n) captured by a sensor (5), wherein the method comprises the following steps:

a. generating a plurality of proposed window specifications (20) and generating the at least one processing specification (30) based on at least one of the proposed window specifications (20);

b. wherein the plurality of proposed window specifications (20) is generated based on a similarity metric adapted for identifying similar pairs of data items (100a, 100b, . . . , 100n) in the stream (10) of data items.

2. The method of claim 1, wherein the similarity metric is adapted for calculating a pair-wise similarity value of at least one pair of data items (100a, 100b, . . . , 100n) in the stream (10) of data items.

3. The method of claim 1, comprising the further step of calculating an accumulated similarity value for the plurality of proposed window specifications (20) and generating the at least one processing specification (30) based on the proposed window specifications (20) with the highest accumulated similarity values.

4. The method of claim 1, comprising the further steps of displaying the plurality of proposed window specifications (20) to a user, selecting at least one of the proposed window specifications (20) by the user and generating the at least one processing specification (30) based on the at least one selected proposed window specifications (20).

5. The method of claim 1, comprising the further steps of selecting a first data item (100a, 100b, . . . , 100n) in the stream (10) and identifying a second data item (100a, 100b, . . . , 100n) in the stream (10) based on the similarity metric.

6. The method of claim 1, comprising the further steps of selecting a third data item (100a, 100b, . . . , 100n) in the stream (10) that is different from the identified second data item and generating a proposed window specification (20) based on the pair of data items formed by the selected first and third data items.

7. The method of claim 1, wherein generating the at least one processing specification (30) comprises generating processing instructions in a stream processing language, preferably XQuery.

8. The method of any of claim 1, wherein the data items (100a, 100b, . . . , 100n) comprise at least one attribute and wherein the similarity metric is adapted for calculating a pair-wise similarity value based on values of the at least one attribute in the at least one pair of data items (100a, 100b, . . . , 100n).

9. The method of claim 1, wherein the pair-wise similarity value is increased, if the values of the at least one attribute in the at least one pair of data items (100a, 100b, . . . , 100n) are equal.

10. The method of claim 8, wherein the pair-wise similarity value is increased, if the at least one attribute is a binary attribute and if the values of the at least one binary attribute in the at least one pair of data items (100a, 100b, . . . , 100n) are complementary.

11. The method of claim 1, comprising the further steps of receiving at least one processing specification (30), deriving at least one stream (10) of data items (100a, 100b, . . . , 100n) based on the at least one processing specification (30) and/or displaying at least one proposed window specification (20) in the stream of data items.

12. A computer program comprising instructions for implementing a method of claim 1.

14. The system (1) of claim 1, wherein the window specification generator (40) is further adapted for calculating an accumulated similarity value for the plurality of proposed window specifications (20) and wherein the processing specification generator (50) is further adapted for generating the at least one processing specification (30) based on the proposed window specifications (20) with the highest accumulated similarity values.

13. A system (1) for generating at least one processing specification (30) for a stream (10) of data items (100a, 100b, . . . , 100n) captured by a sensor (5), wherein the system (1) comprises:

a. a window specification generator (40) adapted for generating a plurality of proposed window specifications (20); and

b. a processing specification generator (50) adapted for generating the at least one processing specification (30) based on at least one of the proposed window specifications (20);

c. wherein the window specification generator (40) is adapted for generating the plurality of proposed window specifications (20) based on a similarity metric adapted for identifying similar pairs of data items (100a, 100b, . . . , 100n) in the stream (10) of data items.

15. The system (1) of claim 13, further comprising a graphical user interface (60) adapted for displaying the plurality of proposed window specifications (20) to a user and for allowing the user to select at least one of the proposed window specifications (20), wherein the processing specification generator (50) is adapted for generating the at least one processing specification (30) based on the at least one selected proposed window specifications (20).