Multimedia Content Popularity: Learning and Recommending a Prediction Method

In 5G networks, Mobile Edge Computing (MEC) has been proposed to enable computation and storage capabilities at the edge of Radio Access Networks. Proactive content caching in MEC is crucial to guarantee users' Quality of Experience thanks to the reduction of traffic latency. Predicting content popularity plays a key role in the effectiveness of proactive caching. In this paper, we propose a generic and flexible recommendation framework which allows recommending suitable learning and prediction algorithms among available ones, in order to predict content popularity. The investigated algorithms are categorized into two main classes: tree-based regressors and recurrent neural networks. Through the study case of YouTube video solicitation profiles, our proposed method, called Imputation-Boosted Collaborative-Filtering based Recommending Prediction Method (IBCF-RPM) shows its effectiveness in the prediction of content popularity for various popularity profiles. By running only 30% of the prediction algorithms, randomly chosen, on a given content profile, the proposed recommending method is able to estimate the accuracy of the other predictors and recommend a well-suited predictor for content popularity.

[1]  Weiwei Xia,et al.  Boosting collaborative filtering based on missing data imputation using item's genre information , 2009, 2009 2nd IEEE International Conference on Computer Science and Information Technology.

[2]  William Remus,et al.  Neural Network Models for Time Series Forecasts , 1996 .

[3]  Yoshua Bengio,et al.  On the Properties of Neural Machine Translation: Encoder–Decoder Approaches , 2014, SSST@EMNLP.

[4]  R. Little.  Missing-Data Adjustments in Large Surveys , 1988 .

[5]  Zachary Chase Lipton.  A Critical Review of Recurrent Neural Networks for Sequence Learning , 2015, ArXiv.

[6]  Gerhard Friedrich,et al.  Recommender Systems - An Introduction , 2010 .

[7]  Li Deng,et al.  Achievements and Challenges of Deep Learning , 2015 .

[8]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[9]  E. S. Gardner,et al.  Forecasting Trends in Time Series , 1985 .

[10]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1995, EuroCOLT.

[11]  F. O. Isinkaye,et al.  Recommendation systems: Principles, methods and evaluation , 2015 .

[12]  Ilyas Alper Karatepe,et al.  Big data caching for networking: moving from cloud to edge , 2016, IEEE Communications Magazine.

[13]  Pascale Minet,et al.  ARMA based popularity prediction for caching in Content Delivery Networks , 2017, 2017 Wireless Days.

[14]  J. Friedman.  Stochastic gradient boosting , 2002 .

[15]  E. Bigdeli,et al.  Comparing accuracy of cosine-based similarity and correlation-based similarity algorithms in tourism recommender systems , 2008, 2008 4th IEEE International Conference on Management of Innovation and Technology.

[16]  G. V. Kass.  An Exploratory Technique for Investigating Large Quantities of Categorical Data , 1980 .

[17]  K. B. Letaief,et al.  A Survey on Mobile Edge Computing: The Communication Perspective , 2017, IEEE Communications Surveys & Tutorials.

[18]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[19]  Apostol Natsev,et al.  YouTube-8M: A Large-Scale Video Classification Benchmark , 2016, ArXiv.

[20]  J. Yosinski,et al.  Time-series Extreme Event Forecasting with Neural Networks at Uber , 2017 .

[21]  Yoshua Bengio,et al.  Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling , 2014, ArXiv.

[22]  Benjamin Letham,et al.  Forecasting at Scale , 2018, PeerJ Prepr..

[23]  Ralph D. Snyder,et al.  Forecasting intraday time series with multiple seasonal cycles using parsimonious seasonal exponential smoothing , 2009 .

[24]  Serge Fdida,et al.  A survey on predicting the popularity of web content , 2014, Journal of Internet Services and Applications.

[25]  Taghi M. Khoshgoftaar,et al.  Imputation-boosted collaborative filtering using machine learning classifiers , 2008, SAC '08.

[26]  Dominique Barth,et al.  Popularity prediction in content delivery networks , 2015, 2015 IEEE 26th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC).

[27]  J. Ross Quinlan,et al.  Induction of Decision Trees , 1986, Machine Learning.

[28]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[29]  Naser Al-Falahy,et al.  Technologies for 5G Networks: Challenges and Opportunities , 2017, IT Professional.

[30]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1997, EuroCOLT.

[31]  Taghi M. Khoshgoftaar,et al.  A Survey of Collaborative Filtering Techniques , 2009, Adv. Artif. Intell..

[32]  Gediminas Adomavicius,et al.  Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions , 2005, IEEE Transactions on Knowledge and Data Engineering.

[33]  Yan Zhang,et al.  Mobile Edge Computing: A Survey , 2018, IEEE Internet of Things Journal.

[34]  C. De Mol,et al.  Optimal Combination of Survey Forecasts , 2012 .