Query processor, query processor elements and a method of establishing such a query processor and query processor elements and a domain processor

• US 20040030685A1
• Filed: 06/16/2003
• Published: 02/12/2004
• Est. Priority Date: 12/14/2000
• Status: Active Grant

• Current Assignee: Kofax Incorporated
• Original Assignee: Kapow APS
• Inventors: Helles, Morten, Hansen, Esben Krag
• Granted Patent: US 7,698,277 B2
• Time in Patent Office: Days
• US Class Current: 707/3
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9538 Presentation of query results

Abstract

The invention relates to a domain processor (DP) comprising: at least one robot modeller (RM); at least one domain modeller (DMR); at least one Query Processor Modeller (QPM), said robot modeller (RM) comprising: means for modelling at least one computer-based robot (R); said at least one robot (R) being adapted for accessing at least one web-based data source (DS); said at least data source (DS) comprising entities comprised in a predefined domain (D); said at least one domain modeller (DMR) comprising: means for modelling at least one domain model (DM); means for establishing at least one extraction model (EM) associated with a chosen domain; means for establishing at least one storage model (STM) associated with said chosen domain, said at least one Query Processor Modeller (QPM) comprising: means for selecting at least two Query Processor elements (QPE) from a set of predefined query processor elements (QPE); means for combining at least two of the selected Query Processor elements (QPE); means for executing said associated query processor elements on at least one computer system (CS); at least one of said query processor elements (QPE) of the associated query processor elements being a Robot query processor Elemet (RQPE) adapted for accessing at least one web-based data source (DS).

Claims

1. Domain processor (DP) comprising

at least one robot modeller (RM)

at least one domain modeller (DMR),

at least one Query Processor Modeller (QPM)

said robot modeller (RM) comprising

means for modelling at least one computer-based robot (R),

said at least one robot (R) being adapted for accessing at least one web-based data source (DS),

said at least one data source (DS) comprising entities comprised in a predefined domain (D),

said at least one domain modeller (DMR) comprising

means for modelling at least one domain model (DM) associated with at least one chosen domain, said domain model (DM) comprising at least one extraction model (EM) and at least one storage model (STM),

means for establishing at least one extraction model (EM) associated with a chosen domain,

means for establishing at least one storage model (STM) associated with said chosen domain,

said at least one Query Processor Modeller (QPM) comprising

means for selecting at least two Query Processor elements (QPE) from

a set of predefined query processor elements (QPE),

means for combining at least two of the selected Query Processor elements (QPE),

means for executing said associated query processor elements on at least one computer system (CS),

at least one of said query processor elements (QPE) of associated query processor elements being a Robot query processor Element (RQPE) adapted for accessing at least one web-based data source (DS).

2. Domain processor (DP) according to claim 1,

wherein the domain processor (DP) comprises at least one query processor maintenance manager (QMM), said at least one query processor maintenance manager (QMM) comprising means for executing at least one query processor (QP) established by the domain processor.

3. Robot modeller (RM) comprising

means for modelling at least one computer-based robot (R),

said at least one robot (R) being adapted for accessing at least one web-based data source (DS),

said at least one data source (DS) comprising entities comprised in a predefined domain (D).

4. Domain modeller (DMR) comprising

means for modelling at least one domain model (DM) associated with at least one chosen domain, said domain model (DM) comprising at least one extraction model (EM) and at least one storage model (STM),

means for establishing at least one extraction model (EM) associated with a chosen domain,

means for establishing at least one storage model (STM) associated with said chosen domain,

5. Domain modeller (DMR) according to claim 4, wherein

said domain modeller comprises means for establishing reference mapping between extracted data obtained according to said extraction model (EM) and a conceptual representation of said data.

6. Domain modeller (DMR) according to claim 4 or 5, wherein

said reference mapping defines a set of reference entities describing a number of entities (E), said entities having attributes.

7. Domain modeller (DMR) according to claim 4 to 6, wherein

said domain modeller (DMR) comprises means for establishing at least one language domain dictionary (LDD).

8. Domain modeller (DMR) according to claims 4-7, wherein

said at least one language domain dictionary (LDD) maps the language of the extracted entities into the general language of the query processor (QP).

9. Domain modeller (DMR) according to claims 4-6, wherein

said domain modeller (DMR) comprises means for establishing a set of reference recognition patterns.

10. Query Processor Modeller (QPM) comprising

means for selecting at least two Query Processor elements (QPE) from a set of predefined query processor elements (QPE),

means for combining at least two of the selected Query Processor elements (QPE),

means for executing said associated query processor elements on at least one computer system (CS),

at least one of said query processor elements (QPE) of the associated query processor elements being a Robot query processor Element (RQPE) adapted for accessing at least one web-based data source (DS).

11. Query Processor Modeller (QPM) according to claim 10,

wherein the Query Processor Modeller comprises a graphical user interface (GUI) in the form of a visual programming tool.

12. Query Processor Modeller (QPM) according to claim 10 or 11

wherein said set of query processor elements (QPE) comprises at least two different types of query processor elements,

at least one type being a robot query processor element (RQPE) and at least one type being a trigger query processor element (TQPE).

13. Query processor maintenance manager (QMM) comprising

means for executing at least one query processor (QP) established by the domain processor.

14. Query processor maintenance manager (QMM) according to claim 13, wherein said maintenance manager (QMM) comprises means for monitoring the state of at least one query processor element (QPE) or the performance of at least one query processor element (QPE).

15. Query processor maintenance manager (QMM) according to claim 13 or 14, wherein said domain processor maintenance manager (QMM) comprises means for evaluating the data flow between query processor elements (QPE) of a query processor path.

16. Query processor maintenance manager (QMM) according to claims 13-15, wherein said domain processor maintenance manager (QMM) comprises means for running and visual monitoring of the individual modules of a query processor.

17. Query processor maintenance manager (QMM) according to claims 13-16, wherein said domain processor maintenance manager (QMM) comprises means for running and visual monitoring of a query processor (QP) on element basis.

18. Web-robot

said robot comprising means for extracting information from web-based data sources (DS) in dependency of at least one extraction model (EM), said at least one extraction model comprising reference data structures defining entities and/or entity structures of data sources in a domain.

19. Web-robot according to claim 18,

said robot comprising at least one exchangeable plug-in, said plug-in comprising retrieving routines adapted for reading knowledge stored in said extraction model, said knowledge preferably being domain-specific.

20. Web-robot according to claim 18 or 19, wherein

said plug-in defines a reference mapping between extracted data obtained according to said extraction model (EM) and a conceptual representation of said data.

21. Web-robot according to claims 18-20, wherein said extraction model (EM) is shared between at least two robots.

22. Query processor (QP),

said query processor (QP) comprising a set of web-based data sources (DS), wherein at least two of said data sources (DS) comprise entities according to a domain model (DM),

said query processor (QP) comprising at least three query processor element (QPE),

at least two of said query processor elements (QPE) comprising

a robot (RQPE)

said robot (RQPE) being attached to at least one data source (DS)

said robot comprising means for accessing information from the at least one data source (DS) according to at least one extraction model (EM) associated with said robot (RQPE),

at least one of said query processor elements (QPE) comprising

a trigger (TQPE)

said trigger query processor element (TQPE) comprising means for establishing a query.

23. Query processor (QP) according to claim 22,

wherein at least one of the query processor elements (QPE) comprises a transformer query processor element (TAQPE), a messenger query processor element (MESQPE) or a mediator query processor element (MQPE).

24. Method of establishing at least one query processor (QP),

said query processor (QP) comprising a set of web-based data sources (DS), wherein at least two of said data sources (DS) comprise entities according to a domain model (DM),

said query processor (QP) comprising at least three query processor element (QPE),

at least two of said query processor elements (QPE) comprising

a robot (RQPE),

said robot comprising means for accessing information from the at least one data source (DS) according to at least one extraction model (EM) associated with said robot (RQPE),

at least one of said query processor elements (QPE) comprising

a trigger (TQPE),

said trigger query processor element (TQPE) comprising means for establishing a query,

said method comprising the step of

attaching at least one selected robot query processor element (RQPE) to at least one of the data sources (DS) of the domain,

combining the selected query processor elements into a query processor (QP) by means of a graphical user interface (GUI).

25. Method of establishing at least one query processor (QP) according to claim 24, wherein said graphical user interface (GUI) defines a query processor element path visually on a drag- and drop basis.

26. Method of establishing at least one query processor (QP) according to claim 24 or 25, wherein at least one of the combined query processor elements (QPE) comprises a transformer query processor element (TAQPE), a messenger query processor element (MESQPE) or a mediator query processor element (MQPE).

27. Method of establishing at least one query processor (QP),

said query processor comprising means for accessing data from web-based data sources (DS) of a domain by means at least one user interface (UI)

said method comprising the steps of

selecting a number of query processor element (QPE) at least one of said selected query processor elements (QPE) being a robot query processor element (RQPE),

at least one of said selected query processor elements (QPE) being a trigger query processor element (TQPE),

attaching at least one selected robot query processor element (RQPE) to at least one of the data sources (DS) of the domain,

combining the selected query processor elements into at least one query path defining the data flow in the query processor (QP) between the user interface (UI) and the web-based data sources of the domain, said method comprising a further step of

customizing the at least one individual robot query processor element (RQPE) to the corresponding attached data sources (DS),

customizing at least one of the trigger query processor elements (TRPE) to the query processor (QP).

28. Method of establishing at least one query processor (QP) according to claim 27, wherein at least one of the combined query processor elements (QPE) comprises a transformer query processor element (TAQPE), a messenger query processor element (MESQPE) or a mediator query processor element (MQPE).

29. Method of extracting data from a web-based data source (DS), said method comprising the steps of

identifying and reading attributes and entities of a web-based data source,

converting the read entities into instances of conceptual entities,

verifying whether the read instances correspond with an entity reference base, (ERB).

30. Method of extracting data from a web-based data source according to claim 29, whereby

the read instances are verified to determine whether they correspond with an entity reference base, (ERB) on the basis of entities represented in said conceptual entity-representing format.

31. Method of extracting data from a web-based data source according to claim 29 or 30, whereby the verified instances are modified according to the entity reference base (ERB) by adding information associated with said instances corresponding to said entity reference base.

32. Method of extracting data from a web-based data source according to claims 29-31, said method comprising

correction of the verified instances according to the entity reference base (ERB) by correcting information associated with said instances corresponding to said entity reference base.

33. Method of establishing a query processor,

said query processor being adapted for accessing data on at least two different web-based data sources,

selecting at least two predefined query processor elements (QPE),

combining the selected query processor elements into a desired query processor structure.

34. Method of establishing a query processor according to claim 33,

said at least two predefined query processor elements having different functional characteristics.

35. Method of establishing a query processor according to claims 33 and 34, said method comprising the step of

modifying the selected query processor elements according to the data structure of said web-based data sources.

36. Method of establishing a query processor according to claims 33-35,

wherein said modification of the selected query processor elements comprises at least one plug-in software module, said at least one plug-in defining domain-specific properties of said element.

37. Method of establishing a domain-accessing routine,

said domain comprising a plurality of web-based data sources,

said method comprising the steps of

establishing at least one robot ( ) adapted for retrieving entities stored on said plurality of web-based data sources

establishing at least one reference catalogue,

establishing at least one procedure of verifying the retrieved entities by comparing the read entities with the at least one reference catalogue.

38. Method of establishing a domain-accessing routine according to claim 37,

said method comprising the steps of

establishing at least one storage means

establishing a data-exchanging interface between said at least one robot and at least one storage means.

39. Method of establishing a domain-accessing routine according to claims 37-38, wherein said reference catalogue is a product catalogue.

40. Method of establishing a domain-accessing routine according to claims 37-39, wherein said established procedure of verification comprises modification of the retrieved entities if the verification procedure indicates or proves that a read entity is not valid according to the at least one reference catalogue.

41. Query processor maintenance manager (QMM)

comprising at least one domain processor user interface (DPUI)

said manager (QMM) comprising means for evaluating different modules of at least one query processor (QP),

said means for evaluating different sub-routines of said query processor comprising

means for monitoring the state of at least on query processor element (QPE)

42. Query processor maintenance manager (QMM) according to claim 41, said processor comprising means for automatically forwarding messages to said at least one query processor user interface (DPUI) when certain predefined conditions are met.

43. Query processor maintenance manager (QMM) according to claim 41 or 42 said manager (QMM) comprising means for modifying individual query processor elements/sub-routines.

44. Query processor maintenance manager (QMM) according to claims 41-43, said manager (QMM) comprising means for modifying the query flow in the query processor during execution of the query processor.