Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

SMART NOTIFICATIONS FOR COLLABORATION WORKSPACES

IP.com Disclosure Number: IPCOM000249702D
Publication Date: 2017-Mar-22
Document File: 8 page(s) / 230K

Publishing Venue

The IP.com Prior Art Database

Related People

Amarkanth Ranganamayna: AUTHOR [+3]

Abstract

Techniques presented herein address the problem of information overload in collaboration applications. More specifically, the delivery of information and notifications is made "smart" and contextual so that the user does not become overwhelmed with a continuous barrage of notifications. Using machine-learning algorithms, notifications are automatically categorized into different buckets of importance and delivery.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 29% of the total text.

Copyright 2017 Cisco Systems, Inc. 1

SMART NOTIFICATIONS FOR COLLABORATION WORKSPACES

AUTHORS: Amarkanth Ranganamayna

Aseem Asthana Ashish Chotai

CISCO SYSTEMS, INC.

ABSTRACT

Techniques presented herein address the problem of information overload in

collaboration applications. More specifically, the delivery of information and notifications

is made "smart" and contextual so that the user does not become overwhelmed with a

continuous barrage of notifications. Using machine-learning algorithms, notifications are

automatically categorized into different buckets of importance and delivery.

DETAILED DESCRIPTION

A large number of collaboration applications use the concept of a collaboration

workspace (e.g., conversation, room, etc.). Given the increasing number of team

workspaces that a user will be part of over a period of time, the user can easily become

overwhelmed by the number of notifications the user may receive (e.g., a user/bot may

send a message, make a call, share content, etc.). Also, conventional methods rely upon

static filters that need to be set manually, and as such are very cumbersome. This negatively

impacts the usability of the application.

An architecture using specific machine learning algorithms may solve this problem.

This architecture may include one or more of the following features: (1) feature vector

selection; (2) automatic threshold determination for classifying notification categories; (3)

smart batch notification processing; and (4) tying in calendaring events to provide an

enhanced notification experience.

Given the explosion of workspaces over time, it becomes difficult for anyone,

including executives, to quickly get to the most important workspaces. This architecture

helps resolve this issue, thereby increasing productivity. A machine learning algorithm

dynamically determines the importance of a given collaboration workspace to a user. The

algorithm may be suitably modified to take into consideration additional machine learning

Copyright 2017 Cisco Systems, Inc. 2

features such as calendaring events (e.g., PagerDuty, On-Call, paid time off (PTO), etc.),

user read receipts (e.g., notification reading patterns), time of the day, location, feeds from

other activities, etc. The algorithm may also dynamically modify the weights associated

with each feature as a part of its learning. Users may then easily develop a user interface

(e.g., 'high-priority-workspaces') that shows the workspaces in the order of the priority of

the notifications received. More detailed example use cases are provided below.

1) Key summary of important notifications

David is the vice president of engineering and is busy launching an important new

product in the marketplace. To this end, he has been collaborating with the launch team,

which includes the chief executive officer and key product stakeholders. When David first

checks his workspace (e.g., mobile, desktop, etc.) in the morning, he is notified of

important messages received overnight w...