Browse Prior Art Database

Method and System for Providing Continual Advertisements (Ads) to User based on Past Events

IP.com Disclosure Number: IPCOM000244597D
Publication Date: 2015-Dec-24
Document File: 2 page(s) / 46K

Publishing Venue

The IP.com Prior Art Database

Related People

Raghu Shivaiah: INVENTOR

Abstract

A method and system is disclosed for providing continual advertisements (ads) to user based on past events. The method and system displays ads, in relation to the events that the user might perform as a next step based on the current user interests.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 55% of the total text.

Method and System for Providing Continual Advertisements (Ads) to User based on Past Events

Abstract

A method and system is disclosed for providing continual advertisements (ads) to user based on past events.  The method and system displays ads, in relation to the events that the user might perform as a next step based on the current user interests.

Description

Disclosed is a method and system for providing continual advertisements (ads) based on past events of users.  The method and system displays ads, in relation to the events that the user might perform as a next step based on the current user interests.

In accordance with the method and system, all sub-sub-categories of events are initially stored.  If a product classification lists around 1600 sub-sub-categories and considering 20 important events per sub-sub-category, a total of 32K entries are stored on Ad-Servers.

Initially popular sub-sub-categories like clothes, shoes, and their sequence of events are handcrafted.  Thereafter, campaign performance data is used to remove, or re-align event nodes offline, and updates the event node/context List Cache every day.  The offline system then calculates the context position score, using CTR.  Higher the scores, context positions are moved towards the head of the List, and lower scores the move the context positions towards the tail.  However, if the event node/context remains in the tail for too long, it does no longer belongs to the list, and is terminated.


An event node is described as,
EventNode

{
sub-sub-category   :      LCD TV   
Time to Live           :      36 Hours
Frequency Cap      :      100 Impressions
nextEventNode      :      TVSubscriptions
positionScore         :      score at the current Position
}
Monitoring use...