Method for processing an input particle stream for creating upper levels of KStore

• US 20070214153A1
• Filed: 03/10/2006
• Published: 09/13/2007
• Est. Priority Date: 03/10/2006
• Status: Abandoned Application

• Current Assignee: Unisys Corporation
• Original Assignee: Unisys Corporation
• Inventors: Mazzagatti, Jane
• Application Number:
  US11/373,733
• Publication Number:
  US 20070214153A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2246 Trees, e.g. B+trees

Abstract

A method for completing an incomplete sequence in a KStore having a particle stream, the particle stream having a plurality of input particles including at least one delimiter includes receiving the at least one delimiter within the particle stream to provide a received delimiter and first determining a current K node in accordance with the received delimiter. A match is second determined in accordance with the received delimiter and the current K node to provide a match determination. The KStore is provided with a list of defined delimiters and the second determining includes accessing the list. A determination is made whether the input particle is on the list. The current K node has an adjacent K node and the second determining includes locating the adjacent node in accordance with an asCase list of the current K node to provide a located ascase node.

Claims

1. A method for completing an incomplete sequence in a KStore having a particle stream, said particle stream having a plurality of input particles including at least one delimiter, comprising:

receiving said at least one delimiter within said particle stream to provide a received delimiter;

first determining a current K node in accordance with said received delimiter; and

second determining a match in accordance with said received delimiter and said current K node to provide a match determination.

2. The method for completing an incomplete sequence in a in a KStore of claim 1, wherein said KStore is provided with a list of defined delimiters and said second determining comprises accessing said list of defined delimiters.

3. The method for completing an incomplete sequence in a KStore of claim 2, further comprising determining whether said input particle is on said list of defined delimiters.

4. The method for completing an incomplete sequence in a KStore of claim 1, wherein said current K node has an adjacent K node that is adjacent to said current K node and said second determining comprises locating said adjacent node in accordance with an asCase list of said current K node to provide a located asCase node.

5. The method for completing an incomplete sequence in a KStore of claim 4, wherein said asCase list includes a plurality of asCase nodes further comprising locating a plurality of adjacent nodes in accordance with said ascase list.

6. The method for completing an incomplete sequence in a KStore of claim 5, wherein a learn function of said KStore is disabled further comprising performing no further operations with said received delimiter if no adjacent node of said plurality of adjacent nodes has a Result node that matches said input delimiter.

7. The method for completing an incomplete sequence in a KStore of claim 4, wherein said second determining further comprises determining a Result node of said located asCase node to provide a determined Result node.

8. The method for completing an incomplete sequence in a KStore of claim 7, wherein said second determining further comprises comparing said determined Result node with said received delimiter.

9. The method for completing an incomplete sequence in a KStore of claim 1, wherein said match determination is negative further comprising building a new asCase node of said current K node in accordance with said negative match determination.

10. The method for completing an incomplete sequence in a KStore of claim 1, wherein said match determination is positive to provide a matched node further comprising setting said current K node to said matched node.

11. The method for completing an incomplete sequence in a KStore of claim 9, wherein said new asCase node comprises an end product node.

12. The method for completing an incomplete sequence in a KStore of claim 9, further comprising setting said new asCase node as a new current K node.

13. The method for completing an incomplete sequence in a KStore of claim 9, further comprising incrementing a node count.

14. The method for completing an incomplete sequence in a KStore of claim 1, wherein said match determination is positive further comprising setting said new asCase node as a new current K node.

15. The method for completing an incomplete sequence in a KStore of claim 14, further comprising incrementing a node count.

16. The method for completing an incomplete sequence in a KStore of claim 1, wherein said KStore includes a plurality of KStore levels having respective current K nodes.

17. The method for completing an incomplete sequence in a KStore of claim 16, wherein said first determining comprises accessing current K node data associating said KStore levels of said plurality of KStore levels with their respective current K nodes.

18. The method for completing an incomplete sequence in a KStore of claim 17, wherein said determining of said current K node further comprises determining a KStore level of said plurality of KStore levels in accordance with said received delimiter.

19. The method for completing an incomplete sequence in a KStore of claim 1, further comprising providing delimiter level data.

20. The method for completing an incomplete sequence in a KStore of claim 19, further comprising accessing said delimiter level data in accordance with said received delimiter.

21. The method for completing an incomplete sequence in a KStore of claim 20, wherein said KStore includes a plurality of KStore levels, a plurality of delimiters and a state data structure for representing associations between said KStore levels and said delimiters further comprising determining a current K level in accordance with said state data structure.

22. The method for completing an incomplete sequence in a KStore of claim 1, further comprising determining whether said KStore includes any KStore levels higher than said current KStore level to provide a higher KStore level determination.

23. The method for completing an incomplete sequence in a KStore of claim 22, wherein said KStore is provided with current K node data further comprising determining a further node on a higher KStore level in accordance with said current K node data and setting said further node as a further current K node.

24. The method for completing an incomplete sequence in a KStore of claim 23, wherein said further node comprises a subcomponent node.

25. The method for completing an incomplete sequence in a KStore of claim 1, further comprising setting said current K node to a sequence beginning location.

26. The method for completing an incomplete sequence in a KStore of claim 25, wherein said sequence beginning location comprises a BOT node.

27. The method for completing an incomplete sequence in a KStore of claim 1, wherein said KStore includes a plurality of KStore levels further comprising:

determining a KStore level of said plurality of KStore levels in accordance with said received delimiter to provide a current KStore level; and

determining whether said KStore includes any KStore levels of said plurality of KStore levels higher than said current KStore level to provide a higher KStore level determination.

28. The method for completing an incomplete sequence in a KStore of claim 27, wherein said further match determination is negative.

29. The method for completing an incomplete sequence in a KStore of claim 28, further comprising building a new asCase node of said higher level current K node when said further match determination is negative.

30. The method for completing an incomplete sequence in a KStore of claim 29, wherein said new asCase node comprises a subcomponent node.

31. The method for completing an incomplete sequence in a KStore of claim 30, further comprising setting said subcomponent node as a further current K node.

32. The method for completing an incomplete sequence in a KStore of claim 27, further comprising determining a Result node of an asCase node of a further current K node of said higher KStore level to provide a further determined Result node.

33. The method for completing an incomplete sequence in a KStore of claim 32, further comprising comparing said further determined Result node with a determined end product node at said lower level to provide a matched node.

34. The method for completing an incomplete sequence in a KStore of claim 32, wherein said match determination is positive further comprising setting said asCase node as a new upper level current K node.

35. The method for completing an incomplete sequence in a KStore of claim 34, further comprising incrementing a node count.

36. The method for completing an incomplete sequence in a KStore of claim 1, further comprising receiving no sensor data within said incomplete sequence prior to receiving said at least one delimiter.

37. The method for completing an incomplete sequence in a KStore of claim 27, wherein said match determination is negative and a learn function of said KStore is disabled further comprising performing no further operations with said received delimiter.

38. The method for completing an incomplete sequence in a KStore of claim 1, wherein said KStore includes a plurality of KStore levels having respective current K nodes and said KStore is provided with a state data structure for storing a correspondence between said KStore levels and said current K nodes further comprising first determining said current K node in accordance with said state data structure.