Meta-Learning for Recommendation System with the Multi-Tasking Learning Setting

In the recommendation system, the number of users is often very large, but each user’s interaction behavior is very limited. For a specific user, how to fully utilize the less interactive data to train a personalized recommendation model is a very challenging problem. In this paper, we introduce the framework of multi-tasking learning into the recommendation system and regard the recommendation of different users as different tasks with certain commonalities. To solve this multi-tasking learning problem, we apply the task-agnostic meta-learning(TAML) in the model. The experimental results on the public datasets MovieLens illustrate those recommendation models trained in the multi-tasking learning setting by the meta-learning in this paper outperform the most deep recommendation models and get the best performance.

[1]  Wentong Li,et al.  Estimating conversion rate in display advertising from past erformance data , 2012, KDD.

[2]  John Riedl,et al.  Item-based collaborative filtering recommendation algorithms , 2001, WWW '01.

[3]  Tong Zhang,et al.  Solving large scale linear prediction problems using stochastic gradient descent algorithms , 2004, ICML.

[4]  Yoon Kim,et al.  Convolutional Neural Networks for Sentence Classification , 2014, EMNLP.

[5]  Yunming Ye,et al.  DeepFM: A Factorization-Machine based Neural Network for CTR Prediction , 2017, IJCAI.

[6]  Anh-Phuong Ta,et al.  Factorization machines with follow-the-regularized-leader for CTR prediction in display advertising , 2015, 2015 IEEE International Conference on Big Data (Big Data).

[7]  Pietro Perona,et al.  One-shot learning of object categories , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Fillia Makedon,et al.  Using singular value decomposition approximation for collaborative filtering , 2005, Seventh IEEE International Conference on E-Commerce Technology (CEC'05).

[9]  Jun Wang,et al.  Product-Based Neural Networks for User Response Prediction , 2016, 2016 IEEE 16th International Conference on Data Mining (ICDM).

[10]  Mubarak Shah,et al.  Task Agnostic Meta-Learning for Few-Shot Learning , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[11]  Sergey Levine,et al.  Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks , 2017, ICML.

[12]  Geoffrey E. Hinton,et al.  Rectified Linear Units Improve Restricted Boltzmann Machines , 2010, ICML.

[13]  Gang Fu,et al.  Deep & Cross Network for Ad Click Predictions , 2017, ADKDD@KDD.

[14]  Yoshua Bengio,et al.  Bayesian Model-Agnostic Meta-Learning , 2018, NeurIPS.

[15]  F. Maxwell Harper,et al.  The MovieLens Datasets: History and Context , 2016, TIIS.

[16]  Joshua B. Tenenbaum,et al.  One shot learning of simple visual concepts , 2011, CogSci.

[17]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.