Product placement engine and method
First Claim
1. A computer-implemented method of automatically identifying products for association with a document comprising the steps of:
- providing a products database having product records containing information regarding an associated product;
parsing a document;
generating a word analysis map for the words in the document including the frequency, object type, and function of the words;
determining word scores of the words in the document based on the frequency, object type, and function of the words in the word analysis map;
adjusting the word scores of the words by predetermined weightings and decays corresponding to the use and object type and function of each word in the document, wherein the decays are based upon a decay constant that is decreased for each repetitive occurrence of each of the words;
constructing a keyword query search string using words having the highest adjusted word scores;
searching the product records of the products database to identify products satisfying the keyword query search string;
assigning initial product scores to the identified products based on matches to the keyword query search string;
parsing the product records to identify word matches in each of the product records and the document;
analyzing the product records by comparing the words in the product records to the words of the document to determine if there is a match between a word of a product record and a word in the document;
updating the initial product scores by processing the initial product scores with the adjusted word scores corresponding to the matched word of the product record and the document;
selecting products from the identified products that have the highest updated product scores; and
displaying a product placement unit on a computer display to a user, wherein the product placement unit is based upon the selected products.
4 Assignments
0 Petitions
Accused Products
Abstract
A product placement engine and method for automatically identifying products for association with a document, the engine including a parser, an analysis module adapted to determine word scores and to adjust the word scores of the words by predetermined weightings, a keyword constructor module adapted to construct a keyword query search string using words having the highest word scores, a search engine adapted to search a products database having product records to identify products satisfying the keyword query search string and assign product scores, and a post processing module adapted to identify word matches in each of the product records and the document and update the product score.
-
Citations
42 Claims
-
1. A computer-implemented method of automatically identifying products for association with a document comprising the steps of:
-
providing a products database having product records containing information regarding an associated product; parsing a document; generating a word analysis map for the words in the document including the frequency, object type, and function of the words; determining word scores of the words in the document based on the frequency, object type, and function of the words in the word analysis map; adjusting the word scores of the words by predetermined weightings and decays corresponding to the use and object type and function of each word in the document, wherein the decays are based upon a decay constant that is decreased for each repetitive occurrence of each of the words; constructing a keyword query search string using words having the highest adjusted word scores; searching the product records of the products database to identify products satisfying the keyword query search string; assigning initial product scores to the identified products based on matches to the keyword query search string; parsing the product records to identify word matches in each of the product records and the document; analyzing the product records by comparing the words in the product records to the words of the document to determine if there is a match between a word of a product record and a word in the document; updating the initial product scores by processing the initial product scores with the adjusted word scores corresponding to the matched word of the product record and the document; selecting products from the identified products that have the highest updated product scores; and displaying a product placement unit on a computer display to a user, wherein the product placement unit is based upon the selected products.
-
-
2. A product placement engine for automatically identifying products for association with a document comprising:
-
a parser adapted to parsing a document; an analysis module adapted to generate a word analysis map for the words in the document including the frequency, object type, and function of the words and further adapted to determine word scores of the words in the document based on the frequency, object type, and function of the words in the word analysis map, and to adjust the word scores of the words by predetermined weightings and decays corresponding to the use and object type and function of each word in the document, wherein the decays are based upon a decay constant that is decreased for each repetitive occurrence of each of the words; a keyword constructor module adapted to construct a keyword query search string using words having the highest adjusted word scores; a search engine adapted to search a products database having product records containing information regarding an associated product to identify products satisfying the keyword query search string and assign initial product scores to the identified products based on matches to the keyword query search string; and a post processing module adapted to analyze the product records by comparing the words in the product records to the words of the document to determine if there is a match between a word of a product record and a word in the document and further adapted to identify word matches in each of the product records and the document and update the initial product scores by processing the initial product scores with the adjusted word scores corresponding to the matched word of the product record and the document and provide the updated product scores to a display module to determine a product placement unit to be displayed to a user on a computer display.
-
-
3. A computer readable medium with instructions recorded thereon for automatically identifying products for association with a document, the computer readable medium comprising:
-
instructions for parsing a document; instructions for generating a word analysis map for the words in the document including the frequency, object type, and function of the words; instructions for determining word scores of the words in the document based on the frequency, object type, and function of the words in the word analysis map; instructions for adjusting the word scores of the words by predetermined weightings and decays corresponding to the use and object type and function of each word in the document, wherein the decays are based upon a decay constant that is decreased for each repetitive occurrence of each of the words; instructions for constructing a keyword query search string using words having the highest adjusted word scores; instructions for searching a products database having product records containing information regarding an associated product to identify products satisfying the keyword query search string; instructions for assigning initial product scores to the identified products based on matches to the keyword query search string; instructions for parsing the product records to identify word matches in each of the product records and the document; instructions for analyzing the product records by comparing the words in the product records to the words of the document to determine if there is a match between a word of a product record and a word in the document; instructions for updating the initial product scores by processing the initial product scores with the adjusted word scores corresponding to the matched word of the product record and the document; instructions for selecting products from the identified products that have the highest updated product scores; and instructions for displaying a product placement unit on a computer display to a user, wherein the product placement unit is based upon the selected products.
-
-
4. A product placement system for automatically identifying at least one product for association with a document comprising:
-
a parser adapted to parse a document; an analysis module adapted to generate a word analysis map for the words in the document including the frequency, object type, and function of the words and further adapted to determine word scores of the words in the document based on the frequency, object type, and function of the words in the word analysis map, and to adjust the word scores of the words by predetermined weightings and decays corresponding to the use and object type and function of each word in the document, wherein the decays are based upon a decay constant that is decreased for each repetitive occurrence of each of the words; a keyword constructor module adapted to construct a keyword query search string using words having the highest adjusted word scores; a search engine adapted to search a products database having product records to identify products satisfying the keyword query search string, and to assign initial product scores to the identified products based on matches to the keyword query search string; and a post processing module adapted to analyze the product records by comparing the words in the product records to the words of the document to determine if there is a match between a word of a product record and a word in the document and further adapted to identify word matches in each of the product records and the document and update the initial product scores by processing the initial product scores with the adjusted word scores corresponding to the matched word of the product record and the document and further adapted to select at least one product based at least on the product scores and provide the updated product scores to a display module to determine a product placement unit to be displayed to a user on a computer display. - View Dependent Claims (5)
-
-
6. A computer-implemented method of automatically identifying at least one product having product records for association with a document comprising the steps of:
-
parsing the document; generating a word analysis map for the words in the document including the frequency, object type, and function of the words; determining word scores of the words in the document based on the frequency, object type, and function of the words in the word analysis map; adjusting the word scores of the words by predetermined weightings and decays corresponding to the use and object type and function of each word in the document, wherein the decays are based upon a decay constant that is decreased for each repetitive occurrence of each of the words; constructing a keyword query search string using words having the highest adjusted word scores; searching the product records of products to identify products satisfying the keyword query search string; assigning product scores to the identified products based on matches to the keyword query search string; parsing the product records to identify word matches in each of the product records and the document; analyzing the product records by comparing the words in the product records to the words of the document to determine if there is a match between a word of a product record and a word in the document; updating the initial product scores by processing the initial product scores with the adjusted word scores corresponding to the matched word of the product record and the document; selecting at least one product based at least on the updated product scores; and displaying a product placement unit on a computer display to a user, wherein the product placement unit is based upon the at least one selected product. - View Dependent Claims (7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41)
-
-
42. A computer readable medium with instructions recorded thereon for automatically identifying products for association with a document, the computer readable medium comprising:
-
instructions for parsing a document; instructions for generating a word analysis map for the words in the document including the frequency, object type, and function of the words; instructions for determining word scores of the words in the document based on the frequency, object type, and function of the words in the word analysis map; instructions for adjusting the word scores of the words by predetermined weightings and decays corresponding to the use and object type and function of each word in the document, wherein the decays are based upon a decay constant that is decreased for each repetitive occurrence of each of the words; instructions for constructing a keyword query search string using words having the highest adjusted word scores; instructions for searching for product records to identify products satisfying the keyword query search string; instructions for assigning initial product scores to the identified products based on matches to the keyword query search string; instructions for parsing the product records to identify word matches in each of the product records and the document; instructions for analyzing the product records by comparing the words in the product records to the words of the document to determine if there is a match between a word of a product record and a word in the document; instructions for updating the initial product scores by processing the initial product scores with the adjusted word scores corresponding to the matched word of the product record and the document; instructions for selecting at least one product based at least partially on the product scores ;
anddisplaying a product placement unit on a computer display to a user, wherein the product placement unit is based upon the selected products.
-
Specification