Goal-driven information interpretation and extraction system

• US 5,864,848 A
• Filed: 01/31/1997
• Issued: 01/26/1999
• Est. Priority Date: 01/31/1997
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Finkelstein, Erich S., Horvitz, Eric
• Primary Examiner(s): Amsbury, Wayne
• Assistant Examiner(s): Jung, David Yiuk
• Application Number:
  US08/792,002
• Time in Patent Office: 725 Days
• Field of Search: 707/1-206, 707/500-542, 345/346-349, 600/300, 364/DIG. 1, 364/DIG. 2, 711/152
• US Class Current: 1/1
• CPC Class Codes: G06F 40/274 Converting codes to words; Guess-ahead of partial word inputs
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

A proactive interface management system for ascertaining a computer user'"'"'s goals and automatically extracting untagged data from one or more source objects and applying the data to one or more target objects. The proactive interface management system includes several cooperating program elements including a user interface, a decision engine, an extractor, and an applicator. These cooperating program elements automatically perform the steps of recognizing user gestures, ascertaining user goals, identifying relevant source information in one or more source objects, extracting the relevant source information, and applying the extracted information to one or more target objects.

Claims

1. A method for extracting information from a source object and storing the extracted information in a target object, comprising the steps of:

selecting a source object with untagged information;

selecting a target object that stores data objects extractable from the untagged information of the source object,

interpreting free-text information within the source file;

based on the interpretation, extracting data objects of the type that are storable in the target object from the untagged information of the source object; and

storing the data objects in the target object.

2. The method of claim 1 wherein the step of interpreting free-text information within the source file further comprises the steps of:

accessing a template of data patterns, each data pattern being associated with one or more of the data objects to be stored in the target object;

searching for occurrences of each data pattern within the source object;

tagging each occurrence of each data pattern found within the source object;

assigning each tagged data pattern to at least one of the data objects; and

ranking the data objects in a priority order.

3. The method of claim 2 wherein the step of ranking the data objects in a priority order further comprises the steps of:

applying a Bayesian inference engine to rank the data objects by likelihood of their correctness;

grouping the tagged target data objects into a plurality of clusters; and

ranking the clusters in a priority order.

4. The method of claim 1 wherein the step of selecting a source object further comprises the step of receiving a user command from a user input device selecting the source object.

5. The method of claim 1 wherein the step of selecting a target object further comprises the step of applying an inference engine to select the target object.

6. The method of claim 1 wherein the step of selecting a target object further comprises the step of receiving a user command from a user input device associating the target object with the source object.

7. A method for interpreting and extracting terms from a free-text portion of a source file, comprising the steps of:

retrieving the source file;

retrieving a application-specific template comprising one or mote target fields for accepting assignments, the target fields being associated with data items to be recorded in a target file;

retrieving a template comprising a plurality of expected terms, each expected term being associated with one or more target fields of the application-specific template;

tagging terms, within the free-text portion of the source file, that correspond to one or more of the expected terms;

grouping the tagged terms into a plurality of clusters;

ranking the clusters in a priority order; and

assigning the tagged terms to the target fields in the priority order.

8. The method of claim 7, wherein:

the application-specific template is selected from a plurality of predefined application-specific templates based on a goal inferred from data within the source file;

the template is selected from a plurality of predefined templates based on the goal inferred from data within the source file; and

the target file is selected from a plurality of predefined target files based on the goal inferred from data within the source file.

9. The method of claim 7, further comprising the steps of:

presenting the application-specific template including the target field assignments to a user for review and modification; and

in response to a user command, storing the target field assignments in the target file.

10. The method of claim 9, further comprising the step of:

in response to a user command, presenting a list of alternative assignments for one or more of the target fields.

11. The method of claim 10, further comprising the step of:

in response to a user command, changing the assignment for one or more of the target fields.

12. The method of claim 11, wherein alternative target field assignments are displayed in a pop-up window on a monitor.

13. The method of claim 9, wherein the source file comprises terms in structured fields.

14. The method of claim 13, wherein the source file is an electronic mail file.

16. The method of claim 11, further comprising the step of updating the template to reflect the changed target field assignment.

15. The method 13, wherein the target file is an electronic calendar file.

17. A computer-readable medium storing a program module comprising instructions which, when executed by a computer system comprising a processing unit, a memory storage device, and a user input/output system including a monitor, cause the computer system to interpret and extract terms from a free-text portion of a source file by performing the steps of:

retrieving the source file;

retrieving an application-specific template comprising one or more target fields for accepting assignments, the target fields being associated with data items to be recorded in a target file;

retrieving a template comprising a plurality of expected terms, each expected term being associated with one or more target fields of the application-specific template;

tagging a set of terms, within the free-text portion of the source file, that correspond to one or more of the expected terms;

grouping the tagged terms into a plurality of clusters;

ranking the clusters in a priority order;

assigning the tagged terms to the target fields in the priority order; and

displaying the application-specific template including the selected target field assignments on the monitor for review and modification.

18. A computer-readable medium storing a program module comprising instructions which, when executed by a computer system comprising a processing unit, a memory storage device, and a user input/output system including a monitor, cause the computer system to interpret and extract terms from free-text information by performing the steps of:

displaying a first window on the monitor;

displaying a portion of the free-text information within the first window;

inferring a goal based on the free-text information;

in response to the goal,

selecting a second window from a plurality of application-specific templates, the second window comprising one or more target fields for accepting assignments, the target fields being associated with data items to be recorded in a target file, and

retrieving a template from a plurality of predefined templates, the selected template comprising a plurality of expected terms, each expected term being associated with one or more target fields of the application-specific template;

tagging a set of terms, within the free-text information, that correspond to one or more of the expected terms;

grouping the tagged terms into a plurality of clusters;

ranking the clusters in a priority order;

assigning the tagged terms to the target fields of the second window in the priority order;

displaying the second window on the monitor; and

displaying the assigned terms on the monitor within the target fields of the second window.

19. The computer-readable medium of claim 18, wherein the first and second windows are displayed simultaneously on the monitor.

20. A computer-readable medium storing a program module comprising instructions which, when executed by a computer system comprising a processing unit, a memory storage device, and a user input/output system including a monitor, cause the computer system to perform the steps of:

recognizing a user gesture;

ascertaining a user goal associated with the user gesture;

identifying goal-related source information in one or more source objects;

interpreting free-text information within the source file;

based on the interpretation determining a relevant portion of the source information;

extracting the relevant portion of the source information;

identifying one or more target objects; and

altering the target objects by applying the extracted information to the target objects.

21. The computer-readable medium of claim 20, wherein the step of ascertaining a user goal further comprises the step of inferring a user goal based on a predefined system context and a pattern of user commands received during a current user interface session.

22. The computer-readable medium of claim 20, wherein the instructions further cause the computer to perform the steps of:

before the step of determining a relevant portion of the source information,

presenting on the user input/output system a query whether to proceed with a proposed data extraction,

receiving a user command to proceed with the proposed data extraction, and

retrieving the source objects from the memory storage device.

23. The computer-readable medium of claim 22, wherein the instructions further cause the computer to perform the steps of:

after the step of altering the target objects by applying the extracted information to the target objects,

presenting on the user input/output system a query whether to proceed with a proposed data storage,

receiving a user command to proceed with the proposed data storage, and

storing the altered target objects in the memory storage device.