Geographical location extraction

• US 7,257,570 B2
• Filed: 12/22/2003
• Issued: 08/14/2007
• Est. Priority Date: 11/13/2003
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Patel, Devesh, Riise, Soren, Stipp, Eugene Heinz
• Primary Examiner(s): Rimell; Sam
• Assistant Examiner(s): LEWIS, ALICIA M
• Application Number:
  US10/745,093
• Publication Number:
  US 20050108213A1
• Time in Patent Office: 1,331 Days
• Field of Search: 707/3, 707/100, 707/104.1, 701/200, 701/201
• US Class Current: 707/706
• CPC Class Codes: G06F 16/9537 Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
• CPC Class Codes: Y10S 707/919 Geographic
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching

Abstract

A geographical location extraction method and tool to infer a likely geographical location from one or more search terms entered as a query by a user on a search engine or the like.

Claims

1. A geographical location extraction method to infer a likely geographical location from one or more search terms entered as a query by a user on a search engine or the like, the method comprising utilising a combination of all of the processes of:

word analysis to determine which parts of a search query comprise location names and provide an indicator of the extent to which a given search term or part thereof should be treated as a geographical location;

ranking geographical location names in accordance with the probability that the location names are likely to be searched;

returning a list of probable locations in response to a search term comprising a non-geographical search term associated with a geographical location name;

the user submitting criteria including a locality name, identifying from the criteria the locality name and returning, in dependence on the locality name, a probability that the locality name is associated with a geographical location; and

analysing the order in which words appear in a search query string, in combination with a search in a locations database, to return a likelihood of a search term relating to a list of probable associated locations.

2. The method according to claim 1, wherein the step of word analysis to determine which parts of a search query comprise location names provides an indicator of the extent to which a given search term or part thereof should be treated as a geographical location by assigning a probability that a search term is a geographical location term and another probability that the search term is a non-geographical location term.

3. The method according to claim 2 wherein the method is further refined by calculating a probability split based on a word use frequency and the significance of the word as a geographical location name.

4. The method according to claim 1, wherein ranking geographical location names in accordance with the probability that the location names are likely to be searched comprises ranking geographical location names by country-based statistics and town-based statistics to determine the search likelihood of geographical locations.

5. The method according to claim 4, wherein the probability of a location being the geographical location searched for is weighted by an event occurrence at the geographical location.

6. The method according to claim 1, wherein returning a list of probable locations in response to a search term comprising a non-geographical search term associated with a geographical location name comprises identifying a set of associated search words for a given geographical location name, there being an increased probability of the geographical location name being the term searched for if the non-geographical search term is in the set of associated search words for that geographical location name.

7. The method according to claim 6, wherein the set of associated search words are identified by determining frequency of occurrence of search words in sources known to relate to a particular geographical location.

8. The method according to claim 1 further comprising one or more of the processes of:

categorising search terms in a query into a local activity or a remote activity;

inferring a geographical location from a search query including a point of interest; and

determining a geographical distance between a likely geographical location derived from the search term and a derived geographical location of a user.

9. The method according to claim 8, wherein categorising search terms in a query into a local activity or a remote activity comprises rule-based processing or a look-up operation in a database of search terms and respective probabilities that the search term relates to a local activity (“

near”

) or a remote activity (“

far”

).

10. The method according to claim 8, wherein derived geographical information concerning the location of the user is used in combination with a local activity (“

near”

) or a remote activity (“

far”

) categorisation to determine whether a “

near”

or a “

far”

geographical location to the user location is sought by the search term.

11. The method according to claim 8, wherein inferring a geographical location from a search query including points of interest comprises deriving the geographical location of the user and the points of interest, identifying the geographical distance from the user to the respective points of interest and inferring the geographical location to be the nearest point of interest to the user location.

12. The method of claim 8 comprising a combination of all the processes of claim 8.

13. The method according to claim 1, wherein analysing the order in which words appear in a search query string comprises the further step of identifying the non-appearance of a word in the string and, in combination with a search in a locations database, returning a likelihood of a search term relating to a list of probable associated locations.

14. A computer-readable medium containing a set of instructions for a general purpose computer and adapted to infer a likely geographical location from one or more search terms entered as a query by a user on a search engine or the like, utilising—

a combination of all processes of;

word analysis to determine which parts of a search query comprise location names and provide an indicator of the extent to which a given search term or part thereof should be treated as a geographical location;

ranking geographical location names in accordance with the probability that the location names are likely to be searched;

returning a list of probable locations in response to a search term comprising a non-geographical search term associated with a geographical location name;

the user submitting criteria including a locality name, identifying from the criteria the locality name and returning, in dependence on the locality name, a probability that the locality name is associated with a geographical location; and

analysing the order in which words appear in a search query string, in combination with a search in a locations database, to return a likelihood of a search term relating to a list of probable associated locations.

15. The computer readable medium according to claim 14 further adapted to infer a likely geographical location by using one or more of the processes of inferring a geographical location from a search query including a point of interest;

categorising search terms in a query into a local activity or a remote activity and determining a geographical distance between a likely geographical location derived from the search term and a derived geographical location of a user.