Discriminative clustering

Patent metadata

Field	Value
Publication number	US-10353958-B2
Application number	US-201715463903-A
Country	US
Kind code	B2
Filing date	Mar 20, 2017
Priority date	Mar 20, 2017
Publication date	Jul 16, 2019
Grant date	Jul 16, 2019

Abstract

A visual analytics system includes a memory and a processor. The processor executes a clustering application having an interactive user-interface rendered on a client computer. The clustering application determines a first cluster of data items of a data set, the data items in the first cluster having first attribute values that are similar to each other within a first degree of similarity and determines a second cluster of data items of the data set, the data items in the second cluster having second attribute values that are similar to each other within a second degree of similarity. For visual analytics, the user interface receives a user selection of a third degree of similarity. In response to which, the clustering application determines a third cluster of data items of the data set, the data items in the third cluster being dissimilar to either the first attribute value of the first reference data item or the second attribute value of the second reference data item by at least the third degree of similarity, and visually displays the third cluster of data items on the user interface.

First claim

What is claimed is:

1. A method for visual analytics of multi-attribute data items in a data set, the method comprising:
providing an interactive user interface to a clustering application hosted on a computer;
determining, by the clustering application, a first cluster of data items of a data set, the data items in the first cluster having first attribute values that are similar to each other within a first degree of similarity, the first cluster of data items being represented by a first reference data item;
determining, by the clustering application, a second cluster of data items of the data set, the data items in the second cluster having second attribute values that are similar to each other within a second degree of similarity, the second cluster of data items being represented by a second reference data item;
receiving, on the user interface, a user selection of a third degree of similarity;
determining a third cluster of data items of the data set, the data items in the third cluster being dissimilar to either the first attribute value of the first reference data item or the second attribute value of the second reference data item by at least the third degree of similarity; and
visually displaying the first, second, and third clusters of data items of the data set on the user interface.

2. The method of claim 1 , wherein visually displaying the third cluster of data items of the data set on the user-interface includes representing the data items by pictorial icons on the user interface.

3. The method of claim 1 , wherein determining the first cluster of data items of the data set includes receiving, on the user interface, a user selection of a first reference data item representing the first cluster of data items.

4. The method of claim 1 , further comprising:
providing a query interface for user queries to retrieve data items from the dataset; and
visually displaying the retrieved data items on the user interface for user selection as the first reference data item.

5. The method of claim 1 , wherein determining the first cluster of data items of the data set includes receiving, on the user interface, a user selection of the first degree of similarity.

6. The method of claim 1 , wherein determining the first cluster of data items of the data set includes receiving, on the user interface, a user identification of the first attribute for determining similarity of the data items in the first cluster to each other within a first degree of similarity.

7. The method of claim 1 , wherein the first, second, and third clusters of data items of the data set are visually displayed in a first, second, and third areas on the user interface, respectively, and wherein the first, second, and third areas are three non-overlapping areas on the user interface.

8. A visual analytics system, comprising:
a memory; and
a processor configured to execute a clustering application having an interactive user-interface hosted on a client computer;
wherein the clustering application determines a first cluster of data items of a data set, the data items in the first cluster having first attribute values that are similar to each other within a first degree of similarity, the first cluster of data items being represented by a first reference data item;
wherein the clustering application determines a second cluster of data items of the data set, the data items in the second cluster having second attribute values that are similar to each other within a second degree of similarity, the second cluster of data items being represented by a second reference data item;
wherein the user interface receives a user selection of a third degree of similarity; and
wherein the clustering application determines a third cluster of data items of the data set, the data items in the third cluster being dissimilar to either the first attribute value of the first reference data item or the second attribute value of the second reference data item by at least the third degree of similarity, and visually displays the first, second, and third clusters of data items of the data set on the user interface.

9. The visual analytics system of claim 8 , wherein visually displaying the third cluster of data items of the data set on the user-interface includes representing the data items by pictorial icons on the user interface.

10. The visual analytics system of claim 8 , wherein the user interface includes an UI element for receiving a user selection of a first reference data item representing the first cluster of data items.

11. The visual analytics system of claim 8 , wherein the user interface includes:
a query interface for building user queries to retrieve data items from the dataset; and
a visual display of the retrieved data items for user selection as the first reference data item.

12. The visual analytics system of claim 8 , wherein the user interface includes an UI element for receiving a user selection of the first degree of similarity.

13. The visual analytics system of claim 8 , wherein the user interface includes an UI element for receiving a user identification of the first attribute for determining similarity of the data items in the first cluster to each other within the first degree of similarity.

14. The visual analytics system of claim 13 , wherein the UI element for receiving the user identification of the first attribute is configured to permit user switching of a previous identification of the first attribute to identification of a different attribute for determining similarity of the data items in the first cluster to each other within the first degree of similarity.

15. A non-transitory computer readable medium comprising:
instructions capable of being executed on a processor, which instructions when executed allow a computing device to:
provide an interactive user interface on a client device to a clustering application;
wherein the clustering application determines a first cluster of data items of a data set, the data items in the first cluster having first attribute values that are similar to each other within a first degree of similarity, the first cluster of data items being represented by a first reference data item;
wherein the clustering application determines a second cluster of data items of the data set, the data items in the second cluster having second attribute values that are similar to each other within a second degree of similarity, the second cluster of data items being represented by a second reference data item;
wherein the user interface receives a user selection of a third degree of similarity; and
wherein the clustering application determines a third cluster of data items of the data set, the data items in the third cluster being dissimilar to either the first attribute value of the first reference data item or the second attribute value of the second reference data item by at least the third degree of similarity, and visually displays the first, second, and third clusters of data items of the data set on the user interface.

16. The non-transitory computer readable medium of claim 15 , wherein visually displaying the third cluster of data items of the data set on the user-interface includes representing the data items by pictorial icons on the user interface.

17. The non-transitory computer readable medium of claim 15 , wherein the user interface includes an UI element for receiving a user selection of a first reference data item representing the first cluster of data items.

18. The non-transitory computer readable medium of claim 15 , wherein the

Assignees

• Sap Se

Inventors

• Latzina Markus

Classifications

• G06F16/904 Primary

  Browsing; Visualisation therefor (for navigating the web G06F16/954; browsing optimisation for the web G06F16/957) · CPC title

• G06F16/287

  Visualization; Browsing · CPC title