Browse Prior Art Database

Demographic Based Recommendation System

IP.com Disclosure Number: IPCOM000241098D
Publication Date: 2015-Mar-26
Document File: 3 page(s) / 164K

Publishing Venue

The IP.com Prior Art Database

Abstract

Current methods for content recommendations provide simplified “one size fits all” models, which sacrifice customer satisfaction for simplicity. In these models, whether a user lives in rural Pennsylvania or in downtown center city, the user may get the same recommendation. An obvious pitfall of this methodology is that the majority of users watch content, such as VOD, within a very narrow time window called the peak period between 8 and 11 each night. Providing the user with an irrelevant recommendation during this narrow & critical time window forces the user to spend precious time to search for desired content. If the search

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 54% of the total text.

Page 01 of 3

D E M O G R A P H I C B A S E D R E C O M M E N D A T I O N S Y S T E M Authors: Zhen Zhao and Hussain Abbas

Background

Current methods for content recommendations provide simplified "one size fits all" models, which sacrifice customer satisfaction for simplicity. In these models, whether a user lives in rural Pennsylvania or in downtown center city, the user may get the same recommendation. An obvious pitfall of this methodology is that the majority of users watch content, such as VOD, within a very narrow time window called the peak period between 8 and 11 each night. Providing the user with an irrelevant recommendation during this narrow & critical time window forces the user to spend precious time to search for desired content. If the search takes too long, the user will give up or watch something else. Both outcomes are not desirable.

New Approach

Proposed is a novel system and method capable of providing personalized recommendations based upon social networks that exist within a geographic area, such as a zip-code, with the result being increased customer satisfaction, retention, and profitability. In one implementation, zip-code level demand behavior is combined with US Census data (e.g., Zip Code Census Tract Data), and IRS tax data. Such a combination of data allows a content source to identify distinct social networks that exist within a given zip-code. Each social network receives its own series of recommendations that are relevant to its tastes and preferences. The seamless demographic based social network system that is proposed allows filtering through noise to provide the user with what he truly wants, when he wants it.

The system is different in that it treats the hierarchy of variables collected in a holistic fashion. For some zip-codes, specific variables might be the most important whilst for a neighboring zip-code they could be the least relevant. The system can also be utilized as a staging ground to drive future marketing campaigns and promotional events, and can be used to develop customized product pricing solutions and identify further opportunities for growth & profit.

The system is different in that it can be used forwards and backwards: it can be used to answer questions and it can be used to propose questions that have not yet been asked. Furthermore, it is capable of answering those questions which it itself proposes.

Example Implementation


1. Build a system that takes aggregate user demand and maps it to the zip-code.

2. Use data mining/analysis tools to identify areas of opportunity as well as the casual relationship between the IRS data, census data and the viewership data. One example is described as follows:

P a g e | 1 170...