System and method for scalable semantic stream processing

• US 9,117,167 B2
• Filed: 11/07/2011
• Issued: 08/25/2015
• Est. Priority Date: 11/05/2010
• Status: Active Grant

• Current Assignee: Sirius-Beta Pty Limited
• Original Assignee: Sirius-Beta Corporation
• Inventors: Goranson, Harold Theodore
• Primary Examiner(s): PARK, GRACE A
• Application Number:
  US13/290,439
• Publication Number:
  US 20120117114A1
• Time in Patent Office: 1,387 Days
• Field of Search: 348/143, 707/776
• US Class Current: 1/1
• CPC Class Codes: G06F 16/41 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06N 5/00 Computing arrangements using knowledge-based models
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: H04N 7/181 for receiving images from a plurality of remote sources

Abstract

A system for collaborative analysis from different processes on different data sources. The system uses a unique approach to lightweight temporary data structures in order to allow communication of interim results among processes, and construction of semantically appropriate reports. The data structures are generated in near real time and their lightweight nature supports massive scaling, including many diverse streaming inputs.

Claims

1. A computer processor system for processing data, comprising:

a communications receiver configured to receive a plurality of streams of data from a plurality of respective sensors, wherein the sensors include;

at least one video sensor, at least one satellite sensor, at least one acoustic sensor, and at least one document source;

one or more computer processors configured to;

identify features of the streams that match known patterns based on the streams of data from the plurality of sensors, using methods specific to each type of the sensors, and process the identified features into infon data structures comprising a relation and at least two parameters on which the relation works;

form semantic binary frames (sb-frames) based on the infon data structures, each sb-frame comprising a segment of data from a stream and semantic information that is universal to the streams of data received from each type of sensor that includes natural language expressions describing the segment of data based on an identified purpose of the computer processor system, and each sb-frame being formed using at least one infon data structure in a respective stream of data, a previous infon data structure in the stream of data, and a next infon data structure in the stream of data;

form fiber groups based on the sb-frames, each fiber group comprising a plurality of sb-frames related by logical natural language expressions that relate the segment of data to other segments of data in the stream;

identify situations of interest as control groups based on the fiber groups; and

receive the sb-frames and the control groups as input, and process the sb-frames to identify a presence of situations of interest and produce a situated deduction.

2. The computer system of claim 1, wherein the one or more computer processors configured to process the sb-frames is a wreathing engine and the wreathing engine is configured to synthesize the sb-frames that conform to the control groups to form a series of facts as one or more situated deductions.

3. The computer system of claim 2, wherein the wreathing engine is configured to compare the situated deductions to predetermined threat profiles.

4. The computer system of claim 3, wherein the wreathing engine is configured to trigger an alert if the situated deductions match the predetermined threat profiles.

5. The computer system of claim 1, wherein the semantic information of the sb-frame is data that indicates what the segment of data may represent in a situation.

6. The computer system of claim 1, wherein separate computer processors are configured to perform the steps of forming sb-frames, identifying situations of interest as control groups, and processing the sb-frames to identify a presence of situations of interest.

7. The computer system of claim 1, wherein the control groups are represented as a mathematical expression.

8. The computer system of claim 1, wherein the streams comprise structured data.

9. The computer system of claim 1, further comprising a computer storage medium storing a plurality of known patterns as recognition processor unit references.

10. The computer system of claim 1, wherein the sb-frames are formed using ontological references that are stored on a computer storage medium.

11. The computer system of claim 1, wherein the situations of interest are stored as mission situation references on a computer storage medium.

12. The computer system of claim 1, comprising a plurality of recognition processor units configured to identify features of the streams and form sb-frames, each of the plurality of recognition processor units specially designed for a particular media type and/or recognition need.

13. The computer system of claim 1, further comprising a router that routes the processed sb-frames to an analysis device.

14. The computer system of claim 1, wherein the plurality of streams of data comprise one or more process flows from each of a plurality of business enterprises, and the situations of interest are products.

15. The computer system of claim 1, wherein the control group further comprises information about the reasoning process used to characterize the situations of interest.

16. A method of processing data, comprising:

receiving, using a communications receiver, a plurality of streams of data from a plurality of respective sensors, wherein the sensors include;

at least one video sensor, at least one satellite sensor, at least one acoustic sensor, and at least one document source;

identifying features of the streams that match known patterns based on the streams of data from the plurality of sensors, using methods specific to each type of sensor, and processing the identified features into infon data structures comprising a relation and at least two parameters on which the relation works;

forming semantic binary frames (sb-frames) based on the infon data structures, each sb-frame comprising a segment of data from a stream and semantic information that is universal to the streams of data received from each type of sensor that includes natural language expressions describing the segment of data based on an identified purpose of the computer processor system, and each sb-frame being formed using at least one infon data structure in a respective stream of data, a previous infon data structure in the stream of data, and a next infon data structure in the stream of data;

form fiber groups based on the sb-frames, each fiber group comprising a plurality of sb-frames related by logical natural language expressions that relate the segment of data to other segments of data in the stream;

identifying control groups based on the fiber groups characterizing situations of interest;

synthesizing sb-frames that conform to the control groups to form a series of facts as one or more situated deductions.

17. A device with one or more non-transitory computer-readable mediums storing a program that, when executed, causes a computer processor to:

receive, using a communications receiver, a plurality of streams of data from a plurality of respective sensors, wherein the sensors include;

at least one video sensor, at least one satellite sensor, at least one acoustic sensor, and at least one document source;

identify features of the streams that match known patterns based on the streams of data from the plurality of sensors, using methods specific to each type of sensor, and process the identified features into infon data structure comprising a relation and at least two parameters on which the relation works;

form semantic binary frames (sb-frames) based on the infon data structures, each sb-frame comprising a segment of data from a stream and semantic information that is universal to the streams of data received from each type of sensor that includes natural language expressions describing the segment of data based on an identified purpose of the computer processor system, and each sb-frame being formed using at least one infon data structure in a respective stream of data, a previous infon data structure in the stream of data, and a next infon data structure in the stream of data;

form fiber groups based on the sb-frames, each fiber group comprising a plurality of sb-frames related by logical natural language expressions that relate the segment of data to other segments of data in the stream;

identify control groups based on the fiber groups characterizing situations of interest;

synthesize sb-frames that conform to the control groups to form a series of facts as one or more situated deductions.