Finding the website of a business using the business name
View Patent ↗A system and method are provided for augmenting information on business directory databases. Using the business name contained in a business directory database and Web data mining technology, the website of a business is found and validated, prior to enriching the database entries.
1. A method for augmenting business information in a record of a business directory database, the record corresponding to a business and containing data relating to the business including at least a business name, the method comprising:
conducting a network search using the business name as a search term;
receiving results of the network search including a plurality of URLs;
determining confidence measures, each confidence measure being associated with one of the plurality of URLs, each confidence measure indicating a probability that a URL with which the confidence measure is associated is sponsored by the business;
selecting a target URL based on a comparison of the confidence measures;
conducting, within a domain corresponding to the target URL, a domain search for a match with the data contained in the record of the business directory database;
if no match is found in the domain search, selecting a next target URL based on the comparison of the confidence measures, and repeating the conducting a domain search for a match within a domain corresponding to the next target URL; and
if a match is found in the domain search, verifying the target URL based on the match;
wherein determining confidence measures comprises constructing, for each confidence measure, a finite state machine that is an intersection of a weighted finite state machine encoding the business name and a weighted finite state machine encoding a domain name in the URL with which the confidence measure is associated, and wherein each path of the finite state machine represents a possible string the finite state machine can match, and wherein each path has an associated cost, the cost being inversely related to the likelihood of the path; and
after verifying the target URL, adding the target URL to the corresponding business record in the business directory database.
2. The method of claim 1 , wherein the network search is a search of the World Wide Web.
3. The method of claim 1 , wherein the network search utilizes a commercial search engine.
4. The method of claim 1 , wherein each confidence measure comprises a logistic transform of the cost of the lowest-cost path in the intersection finite state machine.
5. The method of claim 1 , wherein the data relating to the business further includes an address of the business, and
wherein conducting a domain search for a match with the data contained in the record of the business directory database includes searching for a match with the address.
6. The method of claim 1 , further comprising:
adding information retrieved using the target URL to the corresponding business record in the business directory database.
7. A method for augmenting business information in a record of a business directory database, the record corresponding to a business and containing data relating to the business including at least a business name, the method comprising:
conducting a network search using the business name as a search term;
receiving results of the network search including a plurality of URLs;
selecting a subset of the plurality of URLs based on a comparison of confidence measures, each one of the confidence measures being associated with one of the plurality of URLs, each confidence measure indicating a probability that a URL with which the confidence measure is associated is sponsored by the business;
verifying a desired URL of the subset by conducting, within domains corresponding to each URL of the subset of the plurality of URLs, domain searches for data matching data contained in the business directory database record; and
augmenting the record in the business directory database with data found in the domain searches within the domains;
wherein each confidence measure is determined by constructing a finite state machine that is an intersection of a weighted finite state machine encoding the business name and a weighted finite state machine encoding a domain name in the URL with which the confidence measure is associated, and wherein each path of the finite state machine represents a possible string the finite state machine can match, and wherein each path has an associated cost, the cost being inversely related to the likelihood of the path; and
after verifying the target URL, adding the target URL to the corresponding business record in the business directory database.
8. The method of claim 7 , wherein the network search is a search of the World Wide Web.
9. The method of claim 7 , wherein the network search utilizes a commercial search engine.
10. A non-transitory computer-usable medium having computer readable instructions stored thereon for execution by a processor to perform a method for augmenting business information in a record of a business directory database, the record corresponding to a business and containing data relating to the business including at least a business name, the method comprising:
conducting a network search using the business name as a search term;
receiving results of the network search including a plurality of URLs;
determining confidence measures, each confidence measure being associated with one of the plurality of URLs, each confidence measure indicating a probability that a URL with which the confidence measure is associated is sponsored by the business;
selecting a target URL based on a comparison of the confidence measures;
conducting, within a domain corresponding to the target URL, a domain search for a match with the data contained in the record of the business directory database;
if no match is found in the domain search, selecting a next target URL based on the comparison of the confidence measures, and repeating the conducting a domain search for a match within a domain corresponding to the next target URL; and
if a match is found in the domain search, verifying the target URL based on the match;
wherein determining confidence measures comprises constructing, for each confidence measure, a finite state machine that is an intersection of a weighted finite state machine encoding the business name and a weighted finite state machine encoding a domain name in the URL with which the confidence measure is associated, and wherein each path of the finite state machine represents a possible string the finite state machine can match, and wherein each path has an associated cost, the cost being inversely related to the likelihood of the path; and
after verifying the target URL, adding the target URL to the corresponding business record in the business directory database.