Product placement engine and method

• US 7,505,969 B2
• Filed: 08/05/2004
• Issued: 03/17/2009
• Est. Priority Date: 08/05/2003
• Status: Active Grant

• Current Assignee: CBS Interactive Inc. (Paramount Global (f/k/a ViacomCBS Inc.))
• Original Assignee: CBS Interactive Inc. (Paramount Global (f/k/a ViacomCBS Inc.))
• Inventors: Musgrove, Timothy A., Walsh, Robin
• Primary Examiner(s): Vy, Hung T
• Application Number:
  US10/911,733
• Publication Number:
  US 20050065909A1
• Time in Patent Office: 1,685 Days
• Field of Search: 707 2- 5, 707/7, 707/102, 707/101
• US Class Current: 1/1
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/958 Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0256 User search
• CPC Class Codes: G06Q 30/0277 Online advertisement
• CPC Class Codes: G06Q 30/0603 Catalogue ordering
• CPC Class Codes: G06Q 30/0629 for generating comparisons
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation

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.

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.

5. The system of claim 4, wherein the document is a web page.

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.

7. The method of claim 6, further including parsing the product records to identify word matches in each of the product records and the document.

8. The method of claim 7, further including updating the product score by processing the adjusted word scores corresponding to the matched word with the product score of the product for which word matched.

9. The method of claim 8, further including selecting products from the identified products that have the highest updated product scores.

10. The method of claim 6, wherein the document is a web page.

11. The method of claim 6, wherein parsing includes identifying and processing the location of the document to prevent repetitive analysis of the document.

12. The method of claim 6, wherein parsing includes identifying the corresponding constituent part of each word of said parsed document, said constituent parts of the document including at least one of title, sub-titles, main body, headings, meta-tagged keywords, and meta-tagged description.

13. The method of claim 12, wherein parsing includes ignoring advertisements and navigation links in the document.

14. The method of claim 6, wherein parsing includes obtaining information indicating formatting characteristics of said words of the document, said formatting characteristics including at least one of bolding, italicizing, underlining, capitalization, font type and font size.

15. The method of claim 6, wherein adjusting the word scores includes applying one or more weighting constants to the word scores based on the use of each word in the document.

16. The method of claim 15, further including increasing word scores for words identified to be at least one of a title, a sub-title, a headings, a meta-tagged keyword, and a meta-tagged description.

17. The method of claim 15, wherein said one or more weighting constants are applied based on formatting characteristics of said words.

18. The method of claim 17, further including increasing word scores for words having emphasis formatting including at least one of bolding, italicizing, underlining, capitalization, and increased font size.

19. The method of claim 15, further including developing an analysis map of words to facilitate determination of said word scores.

20. The method of claim 15, further including decreasing said weighting constant for each repetitive occurrence of said word based on a decay constant.

21. The method of claim 6, further including ignoring predetermined stop words of the document.

22. The method of claim 6, further including counting variations of a word as an occurrence of the same word.

23. The method of claim 22, wherein the word score for synonyms, hypernyms, and hyponyms of a word are only partially increased.

24. The method of claim 6, further including counting alternative aliases for at least one of a manufacturer, brand, series and model of a product.

25. The method of claim 6, further including normalizing the word scores of the words of the document.

26. The method of claim 6, wherein said keyword query search string is constructed using at least one Boolean logic operand.

27. The method of claim 26, wherein said at least one Boolean logic operand includes an OR operand.

28. The method of claim 6, further including retrieving product records for said identified products satisfying the keyword query search string.

29. The method of claim 28, wherein said product records include at least one of a manufacturer'"'"'s page, a user manual, a marketing document, a merchant'"'"'s page, a review, sales data, and popularity data.

30. The method of claim 28, further including identifying word matches in a product record and the document.

31. The method of claim 30, further including updating the product scores by processing the adjusted word scores corresponding to the matched word with the product score of the product for which word matched.

32. The method of claim 31, wherein the product score is updated by applying a bonus to increase the product score.

33. The method of claim 32, wherein the amount of said bonus for repetitive occurrence of each word is decreased by applying a penalty.

34. The method of claim 32, further including applying a secondary bonus to increase the product score to update the product score, said secondary bonus being based on existence of content comprising at least one of images, specifications, product briefs, reviews and user opinions.

35. The method of claim 31, further including applying a secondary bonus to increase the product score to update the product score, said secondary bonus being based on at least one of preferred brands, preferred resellers, recency of the product, popularity of the product, and usage scenario.

36. The method of claim 31, further including selecting at least one product having the highest updated product score.

37. The method of claim 36, further including associating said at least one product with the document.

38. The method of claim 36, further including applying a minimum product score threshold in selecting said at least one product.

39. The method of claim 6, further including generating a list of a plurality of products having the highest product scores.

40. The method of claim 39, wherein said list of products include links to product pages of a comparison shopping service.

41. The method of claim 40, further including providing at least one of said products identified in said list of products to at least one client as an advertisement.

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 ;

and

displaying a product placement unit on a computer display to a user, wherein the product placement unit is based upon the selected products.