US 11,146,524 B2Granted

Intelligent contextual timelining of notifications

AssigneeInternational Business Machines Corporation

InventorsAaron K. Baughman, Shikhar Kwatra, Raghuveer Prasad Nagar, Sarbajit K. Rakshit

Life of this patent

Estimating the term…

Loading patent…

Abstract

Using a forecasting model, a timeline of future activities is forecasted. Using a constraints analysis model, a set of profile inputs is analyzed to determine a context of a first period in the forecasted timeline. Using a context extraction model, a first notification from a set of notifications is analyzed to determine a context of the first notification. A degree of correspondence is computed between the context of the first period and the context of the first notification. When the degree of correspondence is at least a threshold degree of correspondence, the first notification is scheduled for presentation during the first period regardless of a time of arrival of the first notification.

Claims20

Dependent claims

A method comprising:

forecasting, using a processor and a memory to execute a forecasting model, a forecasted timeline of future activities;
analyzing, using a constraints analysis model, a set of profile inputs to determine a context of a first period in the forecasted timeline;
analyzing, using a context extraction model, a first notification from a set of notifications, the analyzing determining a context of the first notification;
computing a degree of correspondence between the context of the first period and the context of the first notification;
scheduling, responsive to the degree of correspondence being at least a threshold degree of correspondence, the first notification for presentation during the first period regardless of a time of arrival of the first notification
presenting the first notification during the first period;
collecting user feedback about the first notification, the user feedback being indicative of an accuracy of a timelining of the first notification;
and
applying said user feedback to a machine learning model to modify a model selected from the group consisting of the forecasting model, the constraints analysis model and the context extraction model.

6 more elements

A computer usable program product comprising a computer-readable storage device, and program instructions stored on the storage device, the stored program instructions comprising:

program instructions to forecast, using a processor and a memory to execute a forecasting model, a forecasted timeline of future activities;
program instructions to analyze, using a constraints analysis model, a set of profile inputs to determine a context of a first period in the forecasted timeline;
program instructions to analyze, using a context extraction model, a first notification from a set of notifications, the analyzing determining a context of the first notification;
program instructions to compute a degree of correspondence between the context of the first period and the context of the first notification;
program instructions to schedule, responsive to the degree of correspondence being at least a threshold degree of correspondence, the first notification for presentation during the first period regardless of a time of arrival of the first notification
program instructions to present the first notification during the first period;
program instructions to collect user feedback about the first notification, the user feedback being indicative of an accuracy of a timelining of the first notification;
and
program instructions to apply said user feedback to a machine learning model to modify a model selected from the group consisting of the forecasting model, the constraints analysis model and the context extraction model.

6 more elements

A computer system comprising a processor, a computer-readable memory, and a computer-readable storage device, and program instructions stored on the storage device for execution by the processor via the memory, the stored program instructions comprising:

program instructions to forecast, using the processor and the memory to execute a forecasting model, a forecasted timeline of future activities;
program instructions to analyze, using a constraints analysis model, a set of profile inputs to determine a context of a first period in the forecasted timeline;
program instructions to analyze, using a context extraction model, a first notification from a set of notifications, the analyzing determining a context of the first notification;
program instructions to compute a degree of correspondence between the context of the first period and the context of the first notification;
program instructions to schedule, responsive to the degree of correspondence being at least a threshold degree of correspondence, the first notification for presentation during the first period regardless of a time of arrival of the first notification
program instructions to present the first notification during the first period;
program instructions to collect user feedback about the first notification, the user feedback being indicative of an accuracy of a timelining of the first notification;
and
program instructions to apply said user feedback to a machine learning model to modify a model selected from the group consisting of the forecasting model, the constraints analysis model and the context extraction model.

6 more elements

Dependencies are inferred from the claim text.

Facts

Application: 16/547,269Filed 21 Aug 2019
Published: US 2021/0058357 A125 Feb 2021
Granted: 12 Oct 2021
Term ends: Not estimatedThe source records don’t support an estimate.

Classification8 codes

H04L 51/224
Monitoring or handling of messages: providing notification on incoming messages, e.g. pushed notifications of received messages
G06F 9/542
Interprogram communication: event management; Broadcasting; Multicasting; Notifications
G06N 3/0442
Recurrent networks, e.g. Hopfield networks: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
G06N 3/09
Learning methods: supervised learning