An Study Of The Advertising Inventory Prediction And Allocation Algorithms Of IPTV Business

Acted as a new type of network television,IPTV has come to China for over 10 years since 2005,and it owns more than 41 million users by August 2015 due to the driven factors such as policy,market,technology,and so on.At the same time,the new pattern of business also leads to the era of new media advertising,which is no longer limited to traditional TV,radio,paper media or any other channels of communication.However,traffic prediction and allocation algorithms for the current domestic IPTV advertising has not yet fully mature,many enterprises depend more on human experience in practice.This dissertation mainly studies the advertising inventory prediction,allocation and matching problems on IPTV business.Besides the advertising traffic prediction model,an allocation model is also established,of which the constraints are simplified according to the characteristics of the problem,while taking advantage of the convenience of the sparse matrix.Finally,we verify the feasibility of the algorithm model through a new media company's IPTV case.