Evaluation of nodes

• US 9,830,344 B2
• Filed: 02/12/2015
• Issued: 11/28/2017
• Est. Priority Date: 09/17/2012
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Dutta, Archiman
• Primary Examiner(s): Hasan, Syed
• Assistant Examiner(s): ALLEN, NICHOLAS E
• Application Number:
  US14/620,504
• Publication Number:
  US 20150161187A1
• Time in Patent Office: 1,020 Days
• Field of Search: 707748
• CPC Class Codes: G06F 16/2246 Trees, e.g. B+trees
• CPC Class Codes: G06F 16/285 Clustering or classification

Abstract

Disclosed are various embodiments for assessing the quality of a node that comprises a collection of items containing textual data. The homogeneity of the node can be related to its quality. Highly ranked descriptive terms used in the node are identified and quality score is calculated that provides a measure of the quality of the node. Additionally, a node can be examined for outliers to improve node quality.

Claims

1. A non-transitory computer-readable medium embodying a program that, when executed by at least one computing device, causes the at least one computing device to at least:

identify a plurality of items in a node;

identify a plurality of descriptive terms associated with individual ones of the plurality of items in the node;

identify a subset of the plurality of descriptive terms;

calculate a quality score for the node reflecting a homogeneity of the node, the quality score being based at least in part upon a number of the plurality of items in the node having a description including at least one descriptive term among the subset of the plurality of descriptive terms;

determine whether the quality score for the node meets a specified threshold of homogeneity for the node;

determine an item score for at least one of the plurality of items among the node; and

determine whether the item score meets an item score threshold, the item score threshold reflecting a desired level of homogeneity among the plurality of items in the node.

2. The non-transitory computer-readable medium of claim 1, wherein the program further causes the at least one computing device to:

determine a respective item score for individual ones of the plurality of items in the node;

calculate a mean score for the node based at least in part on the respective item score for at least a subset of the individual ones of the plurality of items in the node;

determine whether the mean score meets the specified threshold of homogeneity for the node; and

identify the node as a high quality node in response to determining that the mean score meets the specified threshold of homogeneity for the node.

3. The non-transitory computer-readable medium of claim 1, wherein the program further causes the at least one computing device to:

in response to a determination that the item score fails to meet the item score threshold, identify the at least one of the plurality of items as an outlier item.

4. The non-transitory computer-readable medium of claim 1, wherein the node comprises at least one of a product category, a genre, or a region.

5. The non-transitory computer-readable medium of claim 1, wherein the node comprises taxonomy data that describes the plurality of items in the node, and the taxonomy data comprises a hierarchical tree.

6. The non-transitory computer-readable medium of claim 1, wherein the program further causes the at least one computing device to:

determine a respective frequency of occurrence of individual ones of the plurality of descriptive terms; and

identify the subset of the plurality of descriptive terms based at least in part on the respective frequency of occurrence of at least a subset of the individual ones of the plurality of descriptive terms.

7. The non-transitory computer-readable medium of claim 1, wherein the plurality of descriptive terms comprise at least one of a bigram or a trigram.

8. The non-transitory computer-readable medium of claim 1, wherein the program further causes the at least one computing device to generate a report indicating at least one of a bad node or a heterogeneous node in response to determining whether the quality score for the node meets the specified threshold of homogeneity for the node.

9. A system comprising:

a data store; and

at least one computing device in communication with the data store, the at least one computing device being configured to at least;

determine a plurality of descriptive terms associated with a plurality of items in a node;

identify a number of highest ranked descriptive terms among the plurality of descriptive terms;

determine a belongingness score for individual ones of the plurality of items based at least in part upon a number of the plurality of items in the node having a description including at least one of the highest ranked descriptive terms;

calculate a quality score for the node based at least in part upon the belongingness score of the individual ones of the plurality of items in the node;

determine whether the quality score for the node meets a specified threshold of homogeneity for the node;

determine whether the belongingness score for an individual item among the plurality of items for the node meets an item score threshold reflecting a desired level of homogeneity among the plurality of items in the node.

10. The system of claim 9, wherein the at least one computing device is further configured to:

identify the individual item as an outlier item in response to a determination that the belongingness score for the individual item fails to meet the specified threshold.

11. The system of claim 9, wherein the quality score for the node indicates a level of homogeneity of the node.

12. The system of claim 11, wherein the at least one computing device is further configured to:

identify the node as a high quality node in response to a determination that the quality score meets the specified threshold of homogeneity for the node.

13. The system of claim 9, wherein the node comprises taxonomy data that describes the plurality of items in the node, and the taxonomy data comprises a hierarchical tree.

14. A method comprising:

identifying, via at least one computing device, a plurality of items in a node;

identifying, via the at least one computing device, a plurality of descriptive terms associated with individual ones of the plurality of items in the node;

identifying, via the at least one computing device, a subset of the plurality of descriptive terms;

determining, via the at least one computing device, which descriptive terms among the subset of the plurality of descriptive terms are associated with the plurality of items in the node;

calculating, via the at least one computing device, a quality score for the node reflecting a homogeneity of the node, the quality score being based at least in part upon a number of the plurality of items in the node having a description including at least one descriptive term among the subset of the plurality of descriptive terms

determining, via the at least one computing device, whether the quality score for the node meets a specified threshold of homogeneity for the node; and

determine whether an item score for at least one of the plurality of items among the node meets an item score threshold, the item score threshold reflecting a desired level of homogeneity among the plurality of items in the node.

15. The method of claim 14, further comprising:

in response to a determination that the item score for the at least one of the plurality of items fails to meet the item score threshold, removing, via the at least one computing device, the at least one of the plurality of items from the node.

16. The method of claim 14, wherein the quality score is based at least in part on a number of times that individual descriptive terms among the subset of the plurality of descriptive terms appear in the plurality of items in the node.

17. The method of claim 14, wherein identifying the subset of the plurality of descriptive terms further comprises:

determining, via the at least one computing device, a respective frequency of occurrence of individual ones of the plurality of descriptive terms; and

identifying, via the at least one computing device, the subset of the plurality of descriptive terms based at least in part on the respective frequency of occurrence of at least a subset of the individual ones of the plurality of descriptive terms.

18. The method of claim 14, wherein the plurality of descriptive terms comprise at least one of a bigram or a trigram.

19. The method of claim 18, further comprising:

generating, via the at least one computing device, a report indicating at least one of a bad node or a heterogeneous node in response to a determination that the quality score for the node fails to meet the specified threshold of homogeneity for the node.

20. The method of claim 14, wherein the node comprises at least one of a product category, a genre, or a region.