Browse Prior Art Database

Method and Apparatus of Notification Context Analysis Manager

IP.com Disclosure Number: IPCOM000244022D
Publication Date: 2015-Nov-05

Publishing Venue

The IP.com Prior Art Database

Abstract

The disclosure introduces a notification context analysis manager to build predictive model to associate user's preference and the pattern of notification, which is identified by context type/keyword/meta-data using context similarity comparison. It will assign a priority to the incoming notification and push it back to the mobile device. Users can configure the priority of the keyword on User Interface (UI). The purpose is to reduce the volume of unimportant and similar notifications and increase the visibility of important notifications.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 01 of 11

Method and Apparatus of Notification Context Analysis Manager Problem Statement and Background:

Overwhelming similar and unimportant notifications


Not every mobile application (app) will replace/merge the existing notifications by the new one
Users may only configure the on/off of notification in app's level
Users do not have control of displaying different kinds of notification in the same app or plurality apps

Core Idea:

Find the pattern of similar notification by analyzing context


Build predictive model by associating notification pattern and user preference Pre-process incoming notification by assigning priority

Collect User's Feedback:

All notifications will be eventually deleted by the user. We can collect user's preference from deleting actions. They can be:
- Press Like, Dislike or neutral area to swipe away
- Swipe away in different direction
- The sequence of deleting
- Whether if users delete without opening,

1


Page 02 of 11

Notification Context Similarity Comparison:

Comparing notifications

Assuming the redundant notifications composed by template and variables

Message = "It's $giftfrom $name. Let's play the game!"

If similarity > threshold (e.g. 70%) => classify as same type

The difference will be variables

2


Page 03 of 11

Analyze Notification Context:

Classify type
Free format or unique notification => Unknown type (neutral)

Identify repeated variables as keywords

A noun exists more than once can be a keyword

Special number or characters can be ignored (e.g. date/time, phone number, email address),

Analyze Notification Metadata:

App name
App classific...