Navigation system with semi-automatic point of interest extraction mechanism and method of operation thereof

• US 9,377,321 B2
• Filed: 11/16/2011
• Issued: 06/28/2016
• Est. Priority Date: 11/16/2011
• Status: Active Grant

• Current Assignee: TeleNav Incorporated
• Original Assignee: TeleNav Incorporated
• Inventors: Diao, Qian, Li, Daqing
• Primary Examiner(s): Beaulieu, Yonel
• Assistant Examiner(s): Kleinman, Lail
• Application Number:
  US13/297,959
• Publication Number:
  US 20130124080A1
• Time in Patent Office: 1,686 Days
• Field of Search: 345/635
• US Class Current: 1/1
• CPC Class Codes: G01C 21/362 received from an external device or application, e.g. PDA, mobile phone or calendar application
• CPC Class Codes: G01C 21/3679 Retrieval, searching and output of POI information, e.g. hotels, restaurants, shops, filling stations, parking facilities G01C21/3611 takes precedence

Abstract

A method of operation of a navigation system includes: extracting a feature from a sample field with a manual label associated with a point of interest, the feature for locating the sample field in a document; generating a rule based on the feature extracted to assign the manual label to a target field; extracting the target field from the document with the manual label assigned by applying the rule to the document; and inserting the target field associated with the point of interest to a location-based guide for displaying on a device.

Claims

1. A method of operation of a navigation system comprising:

extracting a feature from a sample field with a manual label associated with a point of interest, wherein the feature is for locating the sample field in a document;

generating a rule for parsing a high volume of points of interest with a control unit, based on the feature extracted to assign the manual label to a target field;

extracting the target field from the document with the manual label assigned by applying the rule based on the feature extracted from the sample field for training, to the document; and

inserting the target field associated with the point of interest to a location-based guide operated by a device for displaying the inserted target field on the device.

2. The method as claimed in claim 1 wherein generating the rule includes generating the rule with a decision tree.

3. The method as claimed in claim 1 wherein generating the rule includes calculating an information entropy of the sample field.

4. The method as claimed in claim 1 wherein extracting the feature includes extracting the feature having only a binary value.

5. The method as claimed in claim 1 wherein generating the rule includes calculating an information gain of the feature.

6. A method of operation of a navigation system comprising:

extracting a feature from a sample field with a manual label associated with a point of interest, wherein the feature is for locating the sample field in a document;

generating a rule for parsing a high volume of points of interest with a control unit, based on the feature extracted to assign the manual label to a target field;

extracting the target field from the document with the manual label assigned by applying the rule based on the feature extracted from the sample field for training, to the document;

requesting a further sample field when an extraction rate is less than a threshold; and

inserting the target field associated with the point of interest to a location-based guide operated by a device for displaying the inserted target field on the device.

7. The method as claimed in claim 6 further comprising:

receiving a disconfirmation of the rule; and

wherein;

extracting the feature includes requesting the sample field when the disconfirmation of the rule is received.

8. The method as claimed in claim 6 wherein extracting the feature includes extracting the feature being syntactic or semantic.

9. The method as claimed in claim 6 wherein extracting the feature includes extracting the feature derived from regular expression filters.

10. The method as claimed in claim 6 wherein generating the rule includes displaying the rule on a second device.

11. A navigation system comprising:

a control unit;

for extracting a feature from a sample field with a manual label associated with a point of interest, wherein the feature is for locating the sample field in a document;

for generating a rule for paring a high volume of points of interest, based on the feature extracted to assign the manual label to a target field;

for extracting the target field from the document with the manual label assigned by applying the rule based on the feature extracted from the sample field for training, to the document;

for inserting the target field associated with the point of interest to a location-based guide; and

a user interface for displaying the inserted target field on a device.

12. The system as claimed in claim 11 wherein the control unit is for generating the rule with a decision tree.

13. The system as claimed in claim 11 wherein the control unit is for calculating an information entropy of the sample field.

14. The system as claimed in claim 11 wherein the control unit is for extracting the feature having only binary values.

15. The system as claimed in claim 11 wherein control unit is for calculating an information gain of the feature.

16. The system as claimed in claim 11 wherein the control unit is for requesting a further sample field when an extraction rate is less than a threshold.

17. The system as claimed in claim 16 wherein the control unit is for:

receiving a disconfirmation of the rule; and

requesting the sample field when the disconfirmation of the rule is received.

18. The system as claimed in claim 16 wherein the control unit is for extracting the feature being syntactic or semantic.

19. The system as claimed in claim 16 wherein the control unit is for extracting the feature derived from regular expression filters.

20. The system as claimed in claim 16 wherein the control unit is for displaying the rule on a second device.