Method and apparatus for controlling a system

• US 9,897,987 B2
• Filed: 03/19/2013
• Issued: 02/20/2018
• Est. Priority Date: 03/19/2012
• Status: Active Grant

• Current Assignee: British Telecommunications PLC (BT Group PLC)
• Original Assignee: British Telecommunications PLC (BT Group PLC)
• Inventors: Azvine, Behnam, Martin, Trevor Philip
• Primary Examiner(s): Tecklu, Isaac T
• Application Number:
  US14/386,714
• Publication Number:
  US 20150051713A1
• Time in Patent Office: 1,799 Days
• Field of Search: None
• CPC Class Codes: G05B 15/02 electric
• CPC Class Codes: G06F 2218/00 Aspects of pattern recognition specially adapted for signal processing
• CPC Class Codes: G06V 40/20 Movements or behaviour, e.g. gesture recognition recognition of facial expressions G06V40/16
• CPC Class Codes: H04H 60/45 for identifying users

Abstract

A system controller (20, 100, 30, 40) comprises a receiver (20) for receiving data about a system (10, 11) to be controlled and for associating the received data with a time factor; a summarizer (100) for updating a set of summary bins, each of which covers a respective period of time, and each of which stores a summary of the received data having, a time factor which falls within the respective period of time covered by the summary bin; a processor (30) for, processing the summary bins, for example in order to categorize an unknown entity (5) (such as a human user) interacting with the system under control (10, 11) (via a user interface (11) forming part of the system under control) such as into an adult interested in football, etc., and a director (40) for issuing control instructions to the system to be controlled based on the results of the processor (30). The periods of time covered by respective summary bins include overlapping periods of time and periods of time having different durations and are preferably overlapping, fuzzy, multi-scale bins.

Claims

1. A system controller comprising:

an electronic interface configured to receive data about a system to be controlled; and

at least one processor and a memory comprising instructions, the instructions being executable by the at least one processor to at least;

associate the received data with a time factor;

update a set of summary bins, each of which covers a respective period of time, and each of which stores a summary of the received data having a time factor that falls within the respective period of time covered by the summary bin;

process the summary bins; and

issue control instructions to the system to be controlled based on results of the processing of the summary bins;

wherein the periods of time covered by respective summary bins include overlapping periods of time and periods of time having different durations; and

wherein the summary bins are fuzzy bins in the sense that where bins, having the same granularity, cover overlapping periods of time, data associated with a time falling within such an overlap is apportioned between the overlapping summary bins for summarizing purposes in accordance with fuzzy membership rules.

2. The system controller according to claim 1, wherein the updating includes generating and then updating new summary bins, and amalgamating older summary bins into one or more coarser grained summary bins as new data associated with more recent times is received.

3. The system controller according to claim 1, wherein the received data is data resulting from interactions between an unknown entity and the system to be controlled;

wherein the processing of the summary bins includes performing a categorization of the received data in order to identify a category with which to associate the entity, and

wherein the issuance of control instructions includes issuing instructions to the system to be controlled to modify its behavior in order to be tailored to the unknown entity in accordance with the performed categorization.

4. The system controller according to claim 3, wherein the system to be controlled is a social networking platform;

wherein each user of the social networking platform constitutes an unknown entity to be categorized; and

wherein the system is controllable to disable certain functionality for a particular user if the category identified by the processor disagrees with a categorization provided by the user.

5. A summarizer computer node, comprising:

at least one processor and a memory comprising instructions, the instructions being executable by the at least one processor to at least update a set of summary bins, each of which covers a respective period of time, and each of which stores a summary of some received data having a time factor associated therewith that falls within the respective period of time covered by the summary bin;

wherein the periods of time covered by respective summary bins include overlapping periods of time and periods of time having different durations;

wherein the summary bins are fuzzy bins in the sense that where bins, having the same granularity, cover overlapping periods of time, data associated with a time falling within such an overlap is apportioned between the overlapping summary bins for summarizing purposes in accordance with fuzzy membership rules; and

wherein a control instruction is issued to a system to be controlled based on results of processing the summary bins.

6. The summarizer computer node according to claim 5, wherein further instructions are executable to generate and then update new summary bins and to amalgamate older summary bins into one or more coarser grained summary bins as new data associated with more recent times is received.

7. A summarized data reconstruction computing node controllable to provide a reconstruction of summarized data for a specified time interval in connection with a system to be controlled and a system controller, the system controller including an interface configured to receive data about the system to be controlled and processing resources configured to execute a first set of instructions to at least (a) associate the received data with a time factor, (b) update a set of summary bins, each of which covers a respective period of time, and each of which stores a summary of the received data having a time factor that falls within the respective period of time covered by the summary bin, (c) process the summary bins, and (d) issue control instructions to the system to be controlled based on results of the processing of the summary bins, wherein the periods of time covered by respective summary bins include overlapping periods of time and periods of time having different durations, wherein the summary bins are fuzzy bins in the sense that where bins, having the same granularity, cover overlapping periods of time, data associated with a time falling within such an overlap is apportioned between the overlapping summary bins for summarizing purposes in accordance with fuzzy membership rules, and wherein the summarized data reconstruction computing node comprises:

an electronic receiver interface configured to receive an indication of an interval of time of interest for which data is required; and

at least one processor and a memory comprising a second set of instructions, the second set of instructions being executable by the at least one processor to at least;

identify the or each summary bin that is relevant to the received interval of time of interest;

aggregate, in the event that more than one summary bin is identified by the bin identifier, the identified summary bins; and

extract from the summary bin, or from the summary bins after aggregation where the bin identifier identified more than one bin, attribute values.

13. The summarized data reconstruction computing node according to claim 7, wherein the electronic receiver interface is further configured to receive an indication of the attributes of interest.

8. A method of controlling a system, the method comprising:

receiving data about the system being controlled;

associating the received data with a time factor;

generating or updating a set of summary bins and storing the associated values thereof within a suitable data store, each of which summary bins is associated with a respective period of time, and each of which stores a summary of the received data associated with its respective period of time;

processing the summary bins; and

generating and issuing control instructions to the system being controlled based on the results of the processing of the summary bins;

wherein the periods of time associated with respective summary bins include overlapping periods of time and periods of time having different durations; and

wherein the summary bins are fuzzy bins in the sense that where bins, having the same granularity, cover overlapping periods of time, data associated with a time falling within such an overlap is apportioned between the overlapping summary bins for summarizing purposes in accordance with fuzzy membership rules.

9. A non-transient computer readable storage medium comprising processor implementable instructions that, when executed via a computer processor, cause the method of claim 8 to be carried out.

10. The method according to claim 8, wherein the generating or updating includes generating and then updating new summary bins, and amalgamating older summary bins into one or more coarser grained summary bins as new data associated with more recent times is received.

11. The method according to claim 8, wherein the received data is data resulting from interactions between an unknown entity and the system to be controlled;

wherein the processing of the summary bins includes performing a categorization of the received data in order to identify a category with which to associate the entity, and

wherein the issuance of control instructions includes issuing instructions to the system to be controlled to modify its behavior in order to be tailored to the unknown entity in accordance with the performed categorization.

12. The method according to claim 11, wherein the system to be controlled is a social networking platform;

wherein each user of the social networking platform constitutes an unknown entity to be categorized; and

wherein the system is controllable to disable certain functionality for a particular user if the category identified by the processor disagrees with a categorization provided by the user.