Geocoding multi-feature addresses

• US 9,020,941 B1
• Filed: 02/22/2012
• Issued: 04/28/2015
• Est. Priority Date: 01/04/2008
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Khot, Tushar, Gupta, Apurv
• Primary Examiner(s): DANG, THANH HA T
• Application Number:
  US13/402,516
• Time in Patent Office: 1,161 Days
• Field of Search: 707/736, 707/918, 707/919, 707/921
• US Class Current: 707/736
• CPC Class Codes: G06F 16/29 Geographical information databases
• CPC Class Codes: G06F 40/20 Natural language analysis semantic analysis of natural language G06F40/30
• CPC Class Codes: Y10S 707/918 Location
• CPC Class Codes: Y10S 707/919 Geographic
• CPC Class Codes: Y10S 707/921 Spatial

Abstract

A system and method of parsing natural language descriptions of features to determine an approximate location. An embodiment includes splitting the natural language descriptions into components, geocoding each component, and returning the geocode with the highest confidence level. The geocode references a specific location, and this information may be determined by content from a variety of sources. The system may use an assortment of techniques for determining highest confidence level.

Claims

1. A method of geocoding a multi-feature address, comprising:

(a) receiving an address term comprising a region and a feature;

(b) splitting the address term into the region and the feature, the feature including two or more components;

(c) splitting the feature into two or more separate components, wherein each separate components includes at least one component, and each separate component is different from each other separate component, and for each separate component from the feature, determining a respective component geocode representing a location of a component and an associated component confidence score, the determining comprising;

(i) determining at least one geocode based on the region and the component, (ii) determining the respective component geocode representing the location of the component based on the at least one geocode determined in (i), and (iii) determining a component confidence score for the respective component geocode inversely related to a number of geocodes determined in (i) for the component; and

(d) selecting an address geocode representing a location of the address term as the component geocode determined in (c)(ii) according to the corresponding component confidence score determined in (c)(iii),

wherein steps (a)-(d) are each performed on at least one computing device.

2. The method of claim 1, further comprising parsing the address term into the region and the feature based on one or more delimiters in the address term.

3. The method of claim 2, wherein each of the one or more delimiters includes one of a word, a punctuation character, or a blank space.

4. The method of claim 1, the determining (c) further comprising generating geographical coordinate information that includes latitude and longitude values.

5. The method of claim 4, wherein the generating geographical coordinate information comprises:

selecting the geographical coordinate information from a geocode database.

6. The method of claim 4, wherein the generating geographical coordinate information comprises:

selecting the geographical coordinate information from a publicly available user created geocode database.

7. The method of claim 1, wherein the determining (c)(ii) comprises:

determining the respective component geocode as a centroid of the geocodes in (c)(i).

8. A system of geocoding an address, comprising:

an address parser, configured to receive an address term comprising a region and a feature, and split the address term into the region and the feature, the feature including two or more components, and to splitting the feature into two or more separate components, wherein each separate components includes at least one component, and each separate component is different from each other separate component, and

a geocoder configured, for each separate component from the feature, to determine a respective component geocode representing a location of a component and an associated component confidence score by;

(a) determining at least one geocode based on the region and the component,

(b) determining the respective component geocode representing the location of the component based on the at least one geocode determined in (a), and

(c) determining a component confidence score for the respective component geocode inversely related to a number of geocodes determined in (a) for the component; and

a geocode confidence analyzer configured to select an address geocode representing a location of the address term as the component geocode determined by the geocoder in (b) according to the corresponding component confidence score determined by the geocoder in (c),

wherein the address parser, geocoder, and geocode confidence analyzer are implemented on at least one computing device.

9. The system of claim 8, wherein the address parser is configured to parse the address term into the region and the feature based on one or more delimiters in the address term.

10. The system of claim 9, wherein each of the one or more delimiters includes one of a word, a punctuation character, or a blank space.

11. The system of claim 8, wherein the geocoder generates geographical coordinate information that includes latitude and longitude values.

12. The system of claim 11, further comprising:

a geocode database, wherein the geocoder selects the geographical coordinate information from the geocode database.

13. The system of claim 11, further comprising:

a publicly available user created geocode database, wherein the geocoder selects the geographical coordinate information from the geocode database.

14. The system of claim 8, wherein the geocoder determines the component geocode as a centroid of the geocodes determined by the geocoder for the component.