METHOD AND APPARATUS FOR PREDICTING OBJECT PROPERTIES AND EVENTS USING SIMILARITY-BASED INFORMATION RETRIEVAL AND MODEL

• US 20100332474A1
• Filed: 06/25/2010
• Published: 12/30/2010
• Est. Priority Date: 06/25/2009
• Status: Active Grant

• Current Assignee: University of Tennessee Research Foundation (University of Tennessee)
• Original Assignee: University of Tennessee Research Foundation (University of Tennessee)
• Inventors: Wang, Tse-Wei, Icove, David J., Birdwell, J. Douglas, Horn, Sally P.
• Granted Patent: US 8,392,418 B2
• Time in Patent Office: Days
• US Class Current: 707/737
• CPC Class Codes: G06F 16/2228 Indexing structures
• CPC Class Codes: G06F 16/245 Query processing
• CPC Class Codes: G06F 16/2453 Query optimisation
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/3331 Query processing
• CPC Class Codes: G06F 16/355 Class or cluster creation or modification
• CPC Class Codes: G06F 18/2323 based on graph theory, e.g. minimum spanning trees [MST] or graph cuts
• CPC Class Codes: G06F 18/40 Software arrangements specially adapted for pattern recognition, e.g. user interfaces or toolboxes therefor
• CPC Class Codes: G06Q 30/0185 Product, service or business identity fraud
• CPC Class Codes: G06Q 50/265 Personal security, identity or safety
• CPC Class Codes: G06V 10/7635 based on graphs, e.g. graph cuts or spectral clustering
• CPC Class Codes: G06V 10/945 User interactive design; Environments; Toolboxes

Abstract

Method and apparatus for predicting properties of a target object comprise application of a search manager for analyzing parameters of a plurality of databases for a plurality of objects, the databases comprising an electrical, electromagnetic, acoustic spectral database (ESD), a micro-body assemblage database (MAD) and a database of image data whereby the databases store data objects containing identifying features, source information and information on site properties and context including time and frequency varying data. The method comprises application of multivariate statistical analysis and principal component analysis in combination with content-based image retrieval for providing two-dimensional attributes of three dimensional objects, for example, via preferential image segmentation using a tree of shapes and to predict further properties of objects by means of k-means clustering and related methods. By way of example, one of a process failure event, an intrusion event and a fire event and residual objects may be predicted and located and qualified such that, for example, properties of the residual objects may be qualified, for example, via black body radiation and micro-body databases including charcoal assemblages.

Claims

1. A method of determining or localizing a worn component in one of a machine and a process by periodically recording measured data in a machine or process database that supports one of a similarity-based or of a nearest neighbor search during the operation of the machine or process to form a historical record of the operation of the machine or process, receiving an input query comprising recorded measurement data for a current machine or process operating period to the database, and utilizing results returned from the database in response to the query to model component wear status of a component of the machine or process and to predict future behavior of the machine or process, the database being coupled to processor search manager apparatus, the processor search manager apparatus comprising a processor and memory, the method comprising

receiving the input query comprising information representing recorded measurement data from a current machine or process operating period;

utilizing a tree-structured index to retrieve stored information from the historical record that is most similar to said input information;

obtaining component information associated with said retrieved stored information; and

predicting the wear status of at least one component of the machine or process using one of the nearest neighbor and the similarity-based search model for processing said retrieved information comprising the retrieved stored information from the historical record and component information associated with said retrieved stored information.

2. The method of claim 1 wherein said measured data comprise one of electrical, electromagnetic, thermal, and acoustic spectra.

3. The method of claim 1 wherein said measured data comprise one of time series and frequency series data and said model is obtained by solving a convex optimization problem.

4. The method of claim 1 wherein said component information comprises a record of one of a component failure and of component wear measurement data.

5. The method of claim 2 further comprising using principal component analysis to transform said spectra by projecting the spectral information onto a lower dimension subspace and to determine said tree-structured index.

6. The method of claim 1 wherein said measured data are elements of a metric space and said tree-structured index utilizes reference points and a triangle inequality.

7. The method of claim 1 further comprising acquiring new information to add to the historical record, storing said new information to the database, and updating said tree-structured index to permit retrieval of information of said new information similar to information in subsequent queries.

8. The method of claim 7 wherein said updating of said tree-structured index comprises dynamic machine learning.

9. The method of claim 1 further comprising a spectral property database for said machine or process, the spectral property database comprising at least one of thermal, acoustic and electromagnetic properties of components and further comprising predicting one of a fire and an intrusion event.

10. A method of detecting a pattern of criminal activity by recording at least one suspicious transaction in a transaction database, the suspicious transaction database being coupled to processor search manager apparatus, the processor search manager apparatus comprising a processor and memory, the method comprising

testing by said processor of hypotheses representing possible organizational structures of a criminal conspiracy, said suspicious transaction database comprising one of a similarity-based and a nearest neighbor search;

and

applying at least one model to results returned from the database in response to a query, the query requesting objects in said suspicious transaction database most similar to a target, the queried objects most similar to each other, the testing including testing at least one hypothesis and detecting a pattern of criminal activity.

11. A method of predicting a property of a target object via receiving an input query to search manager apparatus coupled to a plurality of databases, the processor search manager apparatus comprising a processor and memory, the method comprising:

generating at least one database of property data for a plurality of objects, an object being described by a vector of attributes;

analyzing via said processor of said search manager apparatus the attributes utilizing multivariate statistical analysis to determine a subspace of an attributes space, the subspace having at least two dimensions;

clustering via said processor the objects into groups of objects where a measure of similarity between objects comprises projecting attribute vectors onto the attributes subspace, the measure being an inverse of the Euclidean distance between points in the attributes subspace;

determining projections via said processor onto a subspace for clustering and indexing objects in said groups of objects in relation to the target object; and

predicting the property of the target object responsive to said clustering and indexing by distance.

12. The method of claim 11 further comprising said processor predicting properties of an object responsive to said clustering of said object into a group of like objects.

13. The method of claim 11 further comprising generating a second database of image data for said plurality of objects.

14. The method of claim 11 wherein said property data comprises one of time series and frequency series data for an object.

15. The method of claim 11 further comprising generating a second database of image data and a third database comprising micro-body assemblage data.

16. The method of claim 11 wherein the attributes subspace is the object clustering and indexing subspace.

17. The method of claim 11 further comprising a spectral data database of one of acoustic and electromagnetic property data, the method further comprising predicting one of a machine failure and a process failure event.

18. The method of claim 11 further comprising a spectral data database of one of acoustic and electromagnetic property data, the method further comprising predicting one of a fire and an intrusion event.

19. A method of predicting object properties of a target object via receiving a query to search manager apparatus coupled to a plurality of databases comprising a core database of property data for objects, the processor search manager apparatus comprising a processor and memory, the method comprising:

determining a set of objects most similar to a target object via said processor search manager apparatus,

applying at least one model to predict a property of the target object,

comparing hypothetical relationships involving the target object and a subset of the set of similar objects, and

determining a probable relationship and a property value of the target object in relation to the set of objects.

20. The method of claim 19 further comprising inferring a property value for a newly acquired object by modeling how a property value varies with other attributes of the newly acquired object.

21. The method of claim 19 wherein the plurality of databases further comprises a database of image data for objects.

22. The method of claim 19 further comprising predicting an occurrence of one of a machine failure and process failure event responsive to said determined probable relationship and property value of the target object.

23. The method of claim 19 further comprising predicting an occurrence of one of a fire and an intrusion event responsive to said determined probable relationship and property value of the target object.

24. The method of claim 19 wherein said at least one model comprises a Bayesian model.

25. A method of predicting object properties of a target object via receiving an input query to search manager apparatus coupled to a plurality of databases for storing objects, the processor search manager apparatus comprising a processor and memory, the method comprising:

searching via said processor search manager apparatus each of at least two different databases for stored objects having an attribute of a type similar to at least one selected attribute of said target object, said selected attribute being different for each database;

retrieving one of a property value and an attribute value in each database associated with said stored objects; and

determining a predicted object property of said target object using said retrieved property or attribute value and a model.

26. The method of claim 25 further comprising

using at least two models and predicted object properties to determine a likely relationship among objects.

27. The method of claim 26 wherein one of the objects is the target object.

28. The method of claim 25 wherein the plurality of databases comprise one of a time series and a frequency series property database and an image database;

the method comprising predicting one of a machine failure event, a process failure event, an intrusion event and a fire event.

29. The method of claim 25 further comprising clustering object data into groups of objects and identifying a group of objects most similar to the target object.

30. The method of claim 25 further comprising predicting an object property of the target object from property data of said group of objects.

31. The method of claim 25 wherein the model comprises a principal component analysis model.

32. The method of claim 25 wherein the model comprises a multivariate data model.

33. The method of claim 25 wherein the model comprises a Bayesian model.

34. The method of claim 25 wherein the model comprises a least squares optimization model.

35. The method of claim 25 wherein the model comprises a maximum likelihood optimization model.

36. The method of claim 25 wherein the model comprises a tree of shapes model.

37. The method of claim 25 wherein the model comprises an Elston and Stewart pedigree model for determining a genetic relationship.

38. The method of claim 25 wherein the model comprises an artificial neural network.

39. The method of claim 25 wherein the model comprises a fuzzy logic model.

40. The method of claim 25 wherein the model comprises a hierarchical data model.

41. A method of predicting an object property of a target object via receiving a query to search manager apparatus coupled to a database, the processor search manager apparatus comprising a processor and memory, the method comprising:

generating at least one database of one of time series data and frequency series data comprising a historical record for each of a plurality of objects, an object being described by a vector of attributes;

analyzing via said processor of said search manager apparatus the attributes utilizing multivariate statistical analysis to determine a subspace of an attributes space, the subspace having at least two dimensions;

clustering via said processor the objects into groups of objects where a measure of similarity between objects comprises projecting attribute vectors onto the attributes subspace, the measure being an inverse of the Euclidean distance between points in the attributes subspace;

determining projections via said processor onto a subspace for clustering and indexing objects in said groups of objects; and

predicting a target object property from said clustering and indexing of groups of objects.

42. The method of claim 41 where the target object is one of a machine and a process and the historical record comprises one of time series data and frequency series data for a process of the machine, the method further comprising a projection of a failure of a component of the machine or process.

43. The method of claim 41 where the one of time series and frequency series data comprises equity price data for a plurality of different equities, the method further comprising an allocation of an equity portfolio.

44. The method of claim 41 further comprising:

generating an impulse response model from the one of time series and frequency series data.

45. The method of claim 44 wherein said impulse response model generation comprises solving a convex optimization problem.

46. A method of storing and organizing information previously stored in one or more databases by using indices to find and retrieve stored objects having attributes similar to the attributes of a target object in order to mine at least one database for knowledge leading to a prediction of a property of the target object, the database being coupled to processor search manager apparatus, the processor search manager apparatus comprising a processor and memory, the method comprising

inputting information describing attributes associated with at least two objects;

storing said information in at least one database and referencing this information using a tree-structured index;

applying multivariate statistical analysis and clustering to said stored information and previously stored information to determine or modify an efficient indexing and search method;

using said efficient indexing and search method to find and retrieve stored objects having attributes similar to the attributes of a target object; and

applying at least one model to the attributes of the retrieved stored objects to obtain a predicted property of the target object.

47. The method of claim 46 wherein the multivariate statistical analysis comprises principal component analysis and the clustering is performed on projections of said stored information and previously stored information to a subspace determined by at least two principal component vectors.

48. The method of claim 46 further comprising applying at least two models corresponding to alternative hypotheses and selecting one of the alternative hypotheses to obtain said predicted property.

49. The method of claim 46 wherein clustering is performed using a k-means algorithm.

50. The method of claim 46 wherein an attribute of some of the stored objects and the target object is one of a time series and a frequency series attribute.

51. The method of claim 50 where an R² correlation coefficient is used to measure a similarity between one of time series and frequency series attributes.

52. The method of claim 46 where an attribute of some of the stored objects and the target object is an image, the model comprising an image segmentation model.