Model-based element configuration in a user interface

Patent metadata

Field	Value
Publication number	US-12547430-B2
Application number	US-202318128106-A
Country	US
Kind code	B2
Filing date	Mar 29, 2023
Priority date	Mar 29, 2023
Publication date	Feb 10, 2026
Grant date	Feb 10, 2026

Abstract

Techniques are described with regard to user interface configuration in a computing environment. An associated computer-implemented method includes initializing an element layout within a set of user interface layers for a certain user based upon random determination, wherein the element layout includes a plurality of elements on which the certain user operates. Responsive to determining that a predefined user history data threshold is exceeded, the method further includes deriving weight metrics for the plurality of elements in association with each of the set of user interface layers based upon user history data, applying at least one layout mode to respective elements among the plurality of elements associated with each of the set of user interface layers based upon the derived weight metrics, and updating the element layout within each of the set of user interface layers for the certain user consequent to applying the at least one layout mode.

First claim

What is claimed is:

1 . A computer-implemented method comprising:
initializing an element layout within a set of user interface layers for a certain user, wherein the element layout is randomly determined responsive to assigning the certain user to a cluster among a plurality of clusters based upon user identification information, and wherein the element layout includes a plurality of elements on which the certain user operates; responsive to determining that a predefined user history data threshold is exceeded for the certain user, deriving weight metrics for the plurality of elements in association with each of the set of user interface layers based upon user history data for the certain user; applying at least one layout mode to respective elements among the plurality of elements associated with each of the set of user interface layers based upon the derived weight metrics; deriving a full space layer weight vector for each element among the plurality of elements based upon any layer among the set of user interface layers in which the certain user operates on the element, wherein the full space layer weight vector is based on the derived weight metrics for the certain user; and updating the element layout within each of the set of user interface layers for the certain user consequent to applying the at least one layout mode.

2 . The computer-implemented method of claim 1 , wherein deriving the weight metrics for the plurality of elements comprises:
calculating an element weight value for each element among the plurality of elements based upon respective operation frequency values for the element for a plurality of slide windows of the user history data, wherein an operation frequency value among the respective operation frequency values represents a number of times the certain user operates on the element; and recalculating the element weight value for each element among the plurality of elements based upon building a time series model incorporating element operation frequency vector data, wherein the element operation frequency vector data is derived by recording an element operation frequency vector for each of the plurality of slide windows that indexes by operation frequency value a predefined number of elements among the plurality of elements.

3 . The computer-implemented method of claim 2 , wherein deriving the weight metrics for the plurality of elements further comprises:
deriving respective layer-specific dot product weight values for each element among the plurality of elements by extracting from the full space layer weight vector portions specific to respective layers in the set of user interface layers.

4 . The computer-implemented method of claim 2 , wherein calculating the element weight value for each element among the plurality of elements comprises:
summing respective products of an operation frequency value for the element and a slide window weight calculated for each of the plurality of slide windows of the user history data, wherein the slide window weight is relatively higher for relatively more recent slide windows among the plurality of slide windows.

5 . The computer-implemented method of claim 2 , wherein building the time series model comprises applying at least one machine learning algorithm to predict at least one future element operation frequency vector for at least one future slide window beyond the plurality of slide windows.

6 . The computer-implemented method of claim 2 , wherein building the time series model comprises training the time series model in conjunction with a long short-term memory recurrent neural network (LSTM-RNN) architecture configured to store at least one time series pattern associated with the element operation frequency vector data.

7 . The computer-implemented method of claim 1 , wherein applying the at least one layout mode to respective elements among the plurality of elements associated with each of the set of user interface layers comprises, for a certain layer in the set of user interface layers:
responsive to determining that a number of elements among the plurality of elements exceeds a predefined element quantity threshold:
applying a focus mode in the certain layer to a subset of elements among the plurality of elements associated with the certain layer having relatively higher layer-specific dot product weight values,
applying a sort mode in the certain layer to the subset of elements in order to arrange the subset of elements by layer-specific dot product weight value, and
applying a disable mode in the certain layer to other elements among the plurality of elements associated with the certain layer.

8 . The computer-implemented method of claim 1 , wherein applying the at least one layout mode to respective elements among the plurality of elements associated with each of the set of user interface layers comprises, for a certain layer in the set of user interface layers:
responsive to determining that a number of elements among the plurality of elements does not exceed a predefined element quantity threshold, calculating for the certain layer a standard deviation value that accounts for all layer-specific dot product weight values; responsive to determining that the calculated standard deviation value exceeds a standard deviation threshold:
applying a focus mode in the certain layer to a subset of elements among the plurality of elements associated with the certain layer having relatively higher layer-specific dot product weight values,
applying a sort mode in the certain layer to the subset of elements in order to arrange the subset of elements by layer-specific dot product weight value, and
applying a disable mode in the certain layer to other elements among the plurality of elements associated with the certain layer.

9 . The computer-implemented method of claim 8 , wherein applying the at least one layout mode to respective elements among the plurality of elements associated with each of the set of user interface layers further comprises, for the certain layer in the set of user interface layers:
responsive to determining that the calculated standard deviation value does not exceed the standard deviation threshold, applying a sort mode in the certain layer to all elements among the plurality of elements associated with the certain layer in order to arrange all elements by layer-specific dot product weight value.

10 . A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to:
initialize an element layout within a set of user interface layers for a certain user, wherein the element layout is randomly determined responsive to assigning the certain user to a cluster among a plurality of clusters based upon user identification information, and wherein the element layout includes a plurality of elements on which the certain user operates; responsive to determining that a predefined user history data threshold is exceeded for the certain user, derive weight metrics for the plurality of elements in association with each of the set of user interface layers based upon user history data for the certain user; apply at least one layout mode to respective elements among the plurality of elements associated with each of the set of user interface layers based upon the derived weight metrics; deriving a full space layer weight vector for each element among the plurality of elements based upon any layer among the set of user interface layers in which the certain user operates on the element, wherein the full space layer weight vector is based on the derived weight metrics

Assignees

• IBM

Inventors

• Wang Jin
• Zhang A Peng
• Gao Lei
• Wu Xian
• Gan Xiang Zhen
• Du Ke

Classifications

• G06F3/0482

  Interaction with lists of selectable items, e.g. menus · CPC title

• G06F9/451 Primary

  Execution arrangements for user interfaces · CPC title