《AM-Bi-LSTM: Adaptive Multi-Modal Bi-LSTM for Sequential Recommendation》

系列论文研读目录

---

---

文章链接

文章题目含义

AM-Bi-LSTM：用于顺序推荐的自适应多模态Bi-LSTM

ABSTRACT

Conventional methods for the early fusion of multi-modal features cannot recognize the relevant modality corresponding to the demand of each user in sequential recommendation. In this paper, we propose the adaptive multi-modal bidirectional long short-term memory network (AM-Bi-LSTM) to recognize the relevant modality for sequential recommendation. Specifically, we construct a new recurrent neural network model that is based on the bidirectional long short-term memory network and obtains multimodal features, including each user’s sequential actions. Our new modality attention module calculates the importance degree of multi-modal features for sequential operations via the late-fusion approach, which results in the method recognizing the relevant modality. In experiments on a multi-modal and sequential dataset including 14,941 clicks constructed from the largest Web service for teachers in Japan, we demonstrate that AM-Bi-LSTM outperforms existing methods in terms of the diversity, explainability, and accuracy ofrecommendation. Specifically, we obtain Recall@10 that is 0.1005 better than that ofexisting early-fusion methods. Moreover, we obtain a value of catalog coverage@10 (representing diversity) that is 0.1710 higher than that for existing methods.传统的多模态特征早期融合方法在顺序推荐中无法识别出与每个用户需求相对应的相关模态。在本文中，我们提出了自适应多模态双向长短期记忆网络（AM-Bi-LSTM）来识别顺序推荐的相关模态。具体来说，我们构建了一个新的递归神经网络模型，该模型基于双向长短期记忆网络，并获得多模态特征，包括每个用户的顺序动作。我们的新的模态注意力模块通过后期融合方法计算多模态特征对顺序操作的重要程度，从而导致该方法识别相关模态。在多模态和顺序数据集上的实验中，包括从日本最大的教师Web服务构建的14，941次点击，我们证明了AM-Bi-LSTM在多样性，可解释性和推荐准确性方面优于现有方法。具体地说，我们得到的Recall@10比现有的早期融合方法好0.1005。此外，我们得到的值的目录coverage@10（代表多样性）是0.1710高于现有的方法。

INDEX TERMS

Sequential recommendation, multi-modal processing, deep learning.
顺序推荐、多模态处理、深度学习。

I. INTRODUCTION

1. In recent years, there has been growing interest in personalized recommendations beyond traditional item-based or user-based recommendations [1], [2], [3], [4]. In particular, sequential recommendation, which captures the evolving preferences of users over time, has gained prominence. Sequential recommendation has many applications in e-commerce, entertainment, and education services.近年来，人们对个性化推荐的兴趣越来越大，超越了传统的基于项目或基于用户的推荐[1]，[2]，[3]，[4]。特别是，顺序推荐，捕捉用户随着时间的推移不断变化的偏好，已经获得了突出地位。顺序推荐在电子商务、娱乐和教育服务中有许多应用。
2. Many researchers attend to the sequential recommendation method, and a survey paper [5] has been published. Conventional methods [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17] use operation histories. Specifically, such methods adopt new techniques such as those of tensor decomposition [6] and recurrent neural networks (RNNs) [10] to estimate users’ dynamic demands. However, these methods cannot recommend items accurately with any/few operation histories. To overcome this problem, researchers have propose