Content Popularity Prediction Methods - A Survey

摘要

Prediction or the forecasting is the way of describing the future event in an efficient manner. Prediction can be used in actuarial science, marketing, child protection, financial services, travel, insurance, telecommunications, retail, mobility, healthcare, pharmaceuticals, capacity planning and other fields. One of the most widely used areas is social networking. So, Social media has an explosive growth of information allowing more users to become content creators and distributors of the web content. So, there exists an online competition, since a few items are getting popular and the remaining items are getting unpopular. Predicting the popularity is very useful in network dimensioning such as caching and replication. In this paper, we present a survey about content popularity methods in various network architectures with various functionalities and present the features that have good forecasting capacity.