注意力机制如今被广泛的应用到各个领域，注意力机制的引入可以使模型的不同部分占据不同的权重从而为最终预测做出不同的贡献。本文基于何向南博士提出的神经协同过滤模型做出改进，在此模型基础上融合长短期记忆网络与广义矩阵分解以捕获用户的长期偏好和短期偏好，同时引入注意力机制，关注用户属性，例如性别，年龄，职业等对推荐效果的影响，为用户属性分配不同的权重，通过用户历史观看序列和用户个人信息进行学习来提升推荐模型的性能。本文基于提出的模型，在MovieLens-1M数据集上进行实验验证，并与其他推荐模型进行对比。实验证明本文提出的融合注意力的长短期记忆网络矩阵分解模型(ALSMF)拥有更好的推荐效果。
The attention mechanism is widely used in various fields. The introduction of the attention mechanism can make different parts of the model occupy different weights to make different contributions to the final prediction. This article is based on the neural collaborative filtering model proposed by Doctor He. On the basis of this model, the long and short-term memory network and generalized matrix decomposition are combined to capture the long-term and short-term preferences of users. At the same time, the attention mechanism is introduced to focus on user’s attributes, such as gender, age, occupation, etc. on the recommendation effect, different weights are assigned to user attributes, and the performance of the recommendation model is improved by learning from the user’s historical viewing sequence and user personal information. This article is based on the proposed model. We use the MovieLens-1M dataset to verify the new model and compare this model, we propose with other recommendation models. Experiments prove that the attention-integrated long and short-term memory network matrix factorization model (ALSMF) proposed in this paper has better recommendation effects.
References
[1]
He, X.N, Liao, L.Z, Zhang, H.W., et al. (2017) Neural Collaborative Filtering. Proceedings of the 26th International Conference on World Wide Web, Pearce, 3-7 April 2017, 173-182. https://doi.org/10.1145/3038912.3052569
Xiao, J., Ye, H., He, X.N., et al. (2017) Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks. Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelli-gence, Melbourne, 19-25 August 2017, 3119-3125. https://doi.org/10.24963/ijcai.2017/435
[4]
Zhou, G.R., Zhu, X.Q., Song, C.R., et al. (2018) Deep Interest Network for Click-Through Rate Prediction. Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, London, 19-23 August 2018, 1059-1068. https://doi.org/10.1145/3219819.3219823
Wu, C.Y., Ahmed, A., Beutel, A., et al. (2017) Recurrent Recommender Networks. Proceedings of the Tenth ACM International Conference on Web Search and Data Mining, Cambridge, 6-10 February 2017, 495-503.
https://doi.org/10.1145/3018661.3018689
