System, a method and a computer program product for performance assessment

• US 8,856,130 B2
• Filed: 02/09/2012
• Issued: 10/07/2014
• Est. Priority Date: 02/09/2012
• Status: Active Grant

• Current Assignee: Kenshoo, Ltd
• Original Assignee: Kenshoo, Ltd
• Inventors: Armon-Kest, Gilad, Siegman, Shahar, Lederhendler, Adina, Benis, Arriel Johan, Zilberbrand, Dorit
• Primary Examiner(s): Shmatov, Alexey
• Assistant Examiner(s): CHEUNG, HUBERT G
• Application Number:
  US13/369,621
• Publication Number:
  US 20130212108A1
• Time in Patent Office: 971 Days
• US Class Current: 707/738
• CPC Class Codes: G06F 16/215 Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A system, a computerized method, and a computer program product for classification of items based on their attributes and on a classification scheme that is defined based on information pertaining to each item of a set of items, and which is indicative of: (a) a quantity of occurrences of the item in a sample; (b) a quantity of successful occurrences of the item in the sample; and (c) at least one attribute of the item with regard to at least one variable out of a set of variables.

Claims

1. A system for classification of advertisements, the system comprising:

a storage apparatus, configured to store information pertaining to items out of a set of items, thereby to yield stored items, the stored information for each given stored item being indicative of;

(a) a total quantity of occurrences of the given stored item in a sample;

(b) a quantity of successful occurrences of the given stored item in the sample; and

(c) one or more attributes of at least one variable characterizing the given stored item, wherein an occurrence of an item is regarded successful if it fulfills a predefined condition associated with a predefined outcome of the occurrence; and

a processor, comprising;

a classification scheme determination module configured to define a classification scheme comprising a plurality of classes for classifying the stored items into classes, defining the classification scheme includes processing the stored information to assign a score to the at least one variable, the assigned score corresponding to a plurality of values corresponding to different attributes out of a plurality of attributes of the at least one variable, each value indicative of a number of successful occurrences of the stored items characterized by the same attribute;

a class management module configured to obtain for each class, out of the plurality of classes, a respective subset of the stored items by applying the defined classification scheme to the stored items, and to determine, with respect to each given class out of the plurality of classes, an outcome estimation value of the given class, the outcome estimation value indicative of the given class performance and based on a value indicative of successful occurrences of the stored items out of the subset of the stored items classified to the given class; and

a performance analysis module configured to classify an analyzing item which does not belong to the set of items by applying the defined classification scheme to attributes characterizing the analyzing item, and to assign to the analyzing item a performance assessment value being a function of the outcome estimation value determined for the given class to which the analyzing item is classified, thereby enabling a selective application of an industrial process, wherein the selective application is responsive to the performance assessment.

2. The system according to claim 1, wherein the defining the classification scheme includes assigning a score to each attribute out of the plurality of attributes of the at least one variable, based on a number of occurrences of stored items characterized by the attribute.

3. The system according to claim 2, wherein the assigning the score further comprises assigning the score to each given attribute out of the plurality of attributes based on a number of successful occurrences of the stored items characterized by the given attribute.

4. The system according to claim 2, wherein the assigning the score further comprises assigning the score to each given attribute out of the plurality of attributes based on a number of unsuccessful occurrences of the stored items characterized by the given attribute.

5. The system according to claim 2, wherein the assigning of the score to each given attribute out of the plurality of attributes comprises computing the score to the given attribute based on:

(a) a number of successful occurrences which are associated with the given attribute in a subset of the sample;

(b) a number of unsuccessful occurrences which are associated with the given attribute in the subset of the sample; and

(c) a total number of occurrences which are associated with the given attribute in the subset of the sample.

6. The system according to claim 1, wherein the items are keywords.

7. The system according to claim 1, wherein each of the items corresponds to a targeted group of people.

8. The system according to claim 1, wherein each occurrence is an impression and each successful occurrence is a click.

9. A computerized method for classification of advertisements, the method comprising:

storing in a non-transitory storage media information pertaining to items of a set of items thereby yielding stored items, the stored information for each given stored item being indicative of;

(a) a total quantity of occurrences of the given stored item in a sample;

(b) a quantity of successful occurrences of the given stored item in the sample; and

(c) one or more attributes of at least one variable characterizing the given stored item, wherein an occurrence of an item is regarded successful if it fulfills a predefined condition associated with a predefined outcome of the occurrence;

defining a classification scheme comprising a plurality of classes and configured for classifying the stored items into classes, wherein the defining the classification scheme comprises processing the stored information to assign a score to the at least one variable, the assigned score corresponding to a plurality of values corresponding to different attributes out of a plurality of attributes of the at least one variable, each value indicative of a number of successful occurrences of the stored items characterized by the same attribute;

applying the defined classification scheme to attributes characterizing the stored items, thereby obtaining for each given class out of the plurality of classes a subset of the stored items in the given class;

determining, with respect to each given class out of the plurality of classes, an outcome estimation value of the given class, the outcome estimation value indicative of the given class performance and based on quantities of successful occurrences of the stored items out of the subset of the stored items classified to the given class;

classifying an analyzing item which does not belong to the set of items by applying the defined classification scheme to attributes characterizing the analyzing item, and

assigning to the analyzing item a performance assessment value being a function of the outcome estimation value determined for the given class to which the analyzing item is classified, thereby enabling a selective application of an industrial process, wherein the selective application is responsive to the performance assessment.

10. The computerized method according to claim 9, wherein the defining the classification scheme includes assigning a score to each given attribute out of the plurality of attributes of the at least one variable, based on a number of occurrences of stored items characterized by the given attribute.

11. The computerized method according to claim 10, wherein the assigning comprises assigning the score to each given attribute out of the plurality of attributes based on a number of successful occurrences of the stored items characterized by the given attribute.

12. The computerized method according to claim 10, wherein the assigning the score comprises assigning the score to each given attribute out of the plurality of attributes based on a number of unsuccessful occurrences of the stored items characterized by the given attribute.

13. The computerized method according to claim 10, wherein the assigning of the score to each given attribute out of the plurality of attributes comprises computing the score to the given attribute based on:

(a) a number of successful occurrences which are associated with the given attribute in a subset of the sample;

(b) a number of unsuccessful occurrences which are associated with the given attribute in the subset of the sample; and

(c) a total number of occurrences which are associated with the given attribute in the subset of the sample.

14. The computerized method according to claim 9, wherein the items are keywords.

15. The computerized method according to claim 14, wherein the storing comprises storing in the non-transitory storage media for each given stored keyword out of a set of stored keywords information in which:

(a) the total quantity of occurrences of the given stored keyword in the sample is indicative of a number of redirections of users which resulted from queries that included the given stored keyword within a sampled time frame;

(b) the quantity of successful occurrences of the given stored keyword in the sample is indicative of a number of conversions resulting from the redirections;

wherein the determining the outcome estimation value comprises determining, with respect to each given class of the plurality of classes a conversion-rate estimation, based on quantities of redirections of stored keywords out of a respective subset of stored keywords in the given class;

wherein the assigning the score comprises assigning to an analyzed keyword which does not belong to the set of stored keywords a conversion-rate assessment which is based on a conversion-rate estimation of the given class corresponding to the analyzing item.

16. The computerized method according to claim 15, further comprising updating an entry that is associated with the analyzed keyword in a bidding database based on the conversion-rate assessment, for facilitating cost reduction in a bidding process that depends on the analyzed keyword.

17. The computerized method according to claim 9, wherein each of the items corresponds to a targeted group of people.

18. The computerized method according to claim 9, wherein each occurrence is an impression and each successful occurrence is a click.

19. The computerized method according to claim 9, wherein each occurrence is a click and each successful occurrence is a conversion.

20. A program storage device readable by machine, tangibly embodying a program of instructions executable by the machine to perform a method for classification of advertisements, comprising the steps of:

storing in a non-transitory storage media information pertaining to items of a set of items, thereby yielding stored items, the stored information for each given stored item being indicative of;

(a) a total quantity of occurrences of the given stored item in a sample;

(b) a quantity of successful occurrences of the given stored item in the sample; and

(c) one or more attributes of at least one variable characterizing the given stored item, wherein an occurrence of an item is regarded successful if it fulfills a predefined condition associated with a predefined outcome of the occurrence;

defining a classification scheme comprising a plurality of classes and configured for classifying the stored items into classes;

wherein the defining the classification scheme comprises;

processing the stored information to assign a score to the at least one variable, the assigned score corresponding to a plurality of values corresponding to different attributes out of a plurality of attributes of the at least one variable, each value indicative of a number of successful occurrences of the stored items characterized by the same attribute;

applying the defined classification scheme to the attributes characterizing the stored items, thereby obtaining for each given class out of the plurality of classes a subset of stored items in the given class;

determining, with respect to each given class out of the plurality of classes, an outcome estimation value of the given class, the outcome estimation value indicative of the given class performance and based on quantities of successful occurrences of the stored items of the subset of stored items classified to the given class; and

classifying an analyzing item which does not belong to the set of items by applying the defined classification scheme to attributes characterizing the analyzing item, and assigning to the analyzing item using a performance assessment value being a function of the outcome estimation value determined for the given class to which the analyzing item is classified, thereby enabling a selective application of an industrial process, wherein the selective application is responsive to the performance assessment.

21. The program storage device according to claim 20, wherein the defining of the classification scheme includes assigning a score to each attribute out of a plurality of attributes of the at least one variable, based on a number of occurrences of stored items characterized by the attribute.

22. The method according to claim 21, wherein the assigning the score comprises assigning the score to each given attribute out of the plurality of attributes based on a number of successful occurrences of the stored items characterized by the given attribute.

23. The program storage device according to claim 21, wherein the assigning the score comprises assigning the score to each given attribute out of the plurality of attributes based on a number of unsuccessful occurrences of the stored items characterized by the given attribute.

24. The program storage device according to claim 21, wherein the assigning of the score to each given attribute out of the plurality of attributes comprises computing the score to the given attribute based on:

(a) a number of successful occurrences which are associated with the given attribute in a subset of the sample;

(b) a number of unsuccessful occurrences which are associated with the given in the subset of the sample; and

(c) a total number of occurrences which are associated with the given attribute in the subset of the sample.

25. The program storage device according to claim 20, wherein the items are keywords.

26. The program storage device according to claim 20, wherein each of the items corresponds to a targeted group of people.

27. The program storage device according to claim 20, wherein each occurrence is an impression and each successful occurrence is a click.

28. The program storage device according to claim 20, wherein each occurrence is a click and each successful occurrence is a conversion.