System and method for self-adapting keyword prediction
View Patent ↗A system and method for implementing keyword prediction for search engine marketing is presented. Historical search data is analyzed to find top search patterns in the form of n-grams. Keywords from historical search data are analyzed to generate n-grams for various values of n. The n-grams can represent trending search term patterns, as opposed to merely finding trending keywords. Thereafter, each of the n-grams is associated with one or more categories and/or sub-categories. The categories and sub-categories are thus associated with one or more n-grams. When a new product is introduced, the new product can be associated with n-grams based on the categories and sub-categories associated with the new product. Advertisements can be bid upon based on the associations. Other embodiments also are disclosed.
1. A system comprising:
one or more processing modules; and
one or more non-transitory memory storage modules storing computing instructions configured to run on the one or more processing modules and perform acts of:
receiving, at a computer processor, a list of historical internet-based product search queries for internet-based products, the list of historical internet-based product search queries having been placed on an e-commerce website of an e-commerce retailer;
generating, using the computer processor, one or more n-grams from a subset of each internet-based product search query in the list of historical internet-based product search queries;
mapping, using the computer processor, each of the one or more n-grams with one or more online terms;
determining, using the computer processor, a category for each term of the one or more online terms;
associating, using the computer processor, a new item with one or more patterns of the one or more n-grams based on an associated category for the new item, the associated category based at least in part on third-party search engine categories of a third-party that is different from the e-commerce retailer;
forming, using the computer processor, a keyword based on a combination of:
(1) one n-gram of the one or more n-grams; and
(2) a common phrase entered into a search engine for each of the internet-based products, the common phrase not comprising the one n-gram;
placing, using the computer processor, a bid on a keyword on a third-party search engine based on the one or more patterns of the one or more n-grams, wherein the third-party search engine is provided by the third-party that is different from the e-commerce retailer; and
facilitating a display of an advertisement for the keyword on the third-party search engine.
2. The system of claim 1 , wherein the computing instructions are further configured to perform an act of:
filtering a determination of the category for each term of the one or more online terms.
3. The system of claim 2 , wherein filtering the determination comprises:
removing duplicates; and
validating the determination to remove a name of a competitor.
4. The system of claim 1 , wherein:
the list of historical internet-based product search queries comprises an indication of a popularity of each internet-based product search query in the list of historical internet-based product search queries, the indication of the popularity comprising a number of times each internet-based product search query in the list of historical internet-based product search queries was searched by a requestor on the e-commerce web site of the e-commerce retailer.
5. The system of claim 4 , wherein:
the subset of each internet-based product search query in the list of historical internet-based product search queries is selected based on the popularity of each internet-based product search query in the list of historical internet-based product search queries.
6. The system of claim 1 wherein:
mapping, using the computer processor, each of the one or more n-grams with the one or more online terms comprises mapping each of the one or more n-grams with 20 online terms, wherein each online term of the 20 online terms is based upon third-party search engine data or a product sold by the e-commerce retailer.
7. The system of claim 1 , wherein:
the category comprises a type of a product sold by the e-commerce retailer; and
the category is divided into one or more sub-categories, the one or more sub-categories comprising a make or model of the product sold by the e-commerce retailer.
8. The system of claim 1 , wherein:
each of the one or more n-grams comprise a series of n consecutive words in an internet-based product query from the list of historical internet-based product search queries; and
n is selected from 1, 2, 3, or 4.
9. The system of claim 1 , wherein:
the computing instructions are further configured to perform an act of:
filtering a determination of the category for each term of the one or more online terms;
filtering the determination comprises:
removing duplicates; and
validating the determination to remove a name of a competitor;
the list of historical internet-based product search queries comprises an indication of a popularity of each internet-based product search query in the list of historical internet-based product search queries, the indication of popularity comprising a number of times each internet-based product search query in the list of historical internet-based product search queries was searched by a requestor on the e-commerce web site of the e-commerce retailer;
the subset of each internet-based product search query in the list of historical internet-based product search queries is selected based on the popularity of each internet-based product search query in the list of historical internet-based product search queries;
mapping, using the computer processor, each of the one or more n-grams with the one or more online terms comprises mapping each of the one or more n-grams with 20 online terms, wherein each online term of the 20 online terms is based upon third-party search engine data or a product sold by the e-commerce retailer;
the category comprises a type of the product sold by the e-commerce retailer;
the category is divided into one or more sub-categories, the one or more sub-categories comprising a make or model of the product sold by the e-commerce retailer;
each of the one or more n-grams comprise a series of n consecutive words in an internet-based product query from the list of historical internet-based product search queries; and
n is selected from 1, 2, 3, or 4.
10. A method comprising:
receiving, at a computer processor, a list of historical internet-based product search queries for internet based products, the list of historical internet-based product search queries having been placed on an e-commerce website of an e-commerce retailer;
generating, using the computer processor, one or more n-grams from a subset of each internet-based product search query in the list of historical internet-based product search queries;
mapping, using the computer processor, each of the one or more n-grams with one or more online terms;
determining, using the computer processor, a category for each term of the one or more online terms;
associating, using the computer processor, a new item with one or more patterns of the one or more n-grams based on an associated category for the new item, the associated category based at least in part on third-party search engine categories of a third-party that is different from the e-commerce retailer;
forming, using the computer processor, a keyword based on a combination of:
(1) one n-gram of the one or more n-grams; and
(2) a common phrase entered into a search engine for each of the internet-based products, the common phrase not comprising the one n-gram; and
placing, using the computer processor, a bid on a keyword on a third-party search engine based on the one or more patterns of the one or more n-grams, wherein the third-party search engine is provided by the third-party that is different from the e-commerce retailer; and
facilitating a display of an advertisement for the keyword on the third-party search engine.
11. The method of claim 10 further comprising:
filtering a determination of the category for each term of the one or more online terms.
12. The method of claim 11 , wherein filtering the determination comprises:
removing duplicates; and
validating the determination to remove a name of a competitor.
13. The method of claim 10 , wherein:
the list of historical internet-based product search queries comprises an indication of a popularity of each internet-based product search query in the list of historical internet-based product search queries, the indication of the popularity comprising a number of times each internet-based product search query in the list of historical internet-based product search queries was searched by a requestor on the e-commerce web site of the e-commerce retailer.
14. The method of claim 13 , wherein:
the subset of each internet-based product search query in the list of historical internet-based product search queries is selected based on the popularity of each internet-based product search query in the list of historical internet-based product search queries.
15. The method of claim 10 , wherein:
mapping, using the computer processor, each of the one or more n-grams with the one or more online terms comprises mapping each of the one or more n-grams with 20 online terms, wherein each online term of the 20 online terms is based upon third-party search engine data or a product sold by the e-commerce retailer.
16. The method of claim 10 , wherein:
the category comprises a type of a product sold by the e-commerce retailer; and
the category is divided into one or more sub-categories, the one or more sub-categories comprising a make or model of the product sold by the e-commerce retailer.
17. The method of claim 10 , wherein:
each of the one or more n-grams comprise a series of n consecutive words in an internet-based product query from the list of historical internet-based product search queries; and
n is selected from 1, 2, 3, or 4.
18. The method of claim 10 , wherein:
the method further comprises:
filtering a determination of the category for each term of the one or more online terms;
filtering the determination comprises:
removing duplicates; and
validating the determination to remove a name of a competitor;
the list of historical internet-based product search queries comprises an indication of a popularity of each internet-based product search query in the list of historical internet-based product search queries, the indication of the popularity comprising a number of times each internet-based product search query in the list of historical internet-based product search queries was searched by a requestor on the e-commerce web site of the e-commerce retailer;
the subset of each internet-based product search query in the list of historical internet-based product search queries is selected based on the popularity of each internet-based product search query in the list of historical internet-based product search queries;
mapping, using the computer processor, each of the one or more n-grams with the one or more online terms comprises mapping each of the one or more n-grams with 20 online terms, wherein each online term of the 20 online terms is based upon third-party search engine data or a product sold by the e-commerce retailer;
the category comprises a type of the product sold by the e-commerce retailer;
the category is divided into one or more sub-categories, the one or more sub-categories comprising a make or model of the product sold by the e-commerce retailer;
each of the one or more n-grams comprise a series of n consecutive words in an internet-based product query from the list of historical internet-based product search queries; and
n is selected from 1, 2, 3, or 4.