Multi-temporal Sequential Recommendation Model Based on the Fused Learning Preferences

Jianxia Chen, Liwei Pan, Shi Dong, Tianci Yu, Liang Xiao, Meihan Yao, Shijie LuoInternational Journal of Computational Intelligence Systems 2023 · DOI

Abstract

Sequential recommendation plays a vital role in modern recommendation systems by capturing dynamic user interaction patterns. However, many deep learning-based models struggle to balance short-term interests with long-term preferences and often fail to fully exploit temporal information within interaction sequences. This paper proposes the MTSR-FLP model, which utilizes an MLP-based state generation module to simultaneously consider long-term and short-term preferences. The framework integrates a global representation learning approach to capture general user trends and a local representation learning approach for historical user-specific information. Furthermore, a novel multi-temporal embedding scheme is developed using multiple kernels for absolute and relative timestamps. Experimental results on five real-world datasets demonstrate that MTSR-FLP significantly outperforms advanced baselines, improving HR@10 by up to 31.10% and NDCG@10 by up to 42.54%.

Introduction

Sequential recommendation (SR) is the engine behind personalized content delivery, predicting a user's next interaction based on their historical sequence. Despite the success of deep learning in this domain, many models suffer from "recency bias," over-prioritizing the most recent clicks while neglecting broader user habits. Additionally, time is often treated as a simple ordinal index rather than a rich contextual signal. MTSR-FLP addresses these gaps by treating time as a multi-dimensional feature and fusing global behavioral trends with local user history.

Methodology

The MTSR-FLP architecture is built upon three core components: multi-temporal embedding, dual-track representation learning (local and global), and an item similarity gating mechanism.

The overall architecture of the MTSR-FLP model, illustrating the integration of local and global feature learning layers with multi-timestamp embeddings.

Multi-temporal Embedding Learning

Rather than a single time vector, MTSR-FLP employs a sophisticated embedding scheme that captures both absolute and relative time patterns.

  • Absolute Embeddings: Combine sequence indices with time-of-day information.
  • Relative Embeddings: Calculate intervals between interactions using sine, exponential, and logarithmic kernels to capture patterns ranging from immediate bursts to long-term periodic habits.
Detailed schematic of the absolute and relative temporal embedding schemes used to encode user-item interactions.

Local and Global Representation

  • Local Representation Learning (LRL): This module uses self-attention to weigh a user's historical interactions, enriched by the multi-temporal embeddings, to capture specific personal preferences.
  • Global Representation Learning (GRL): Utilizing a Location-Based Attention (LBA) layer, this module identifies general item-to-item relationships across the entire user base, providing a "global" context that balances idiosyncratic local patterns.

Item Similarity Gating

A final gating mechanism coordinates the output. It calculates the similarity between a candidate item and the user's recent history to dynamically decide the weight of local versus global preferences for a specific recommendation.

Experimental Results

The researchers evaluated MTSR-FLP against state-of-the-art models like SASRec and GRU4Rec on five diverse datasets: Tmall, Steam, Amazon Games, Amazon Beauty, and Foursquare.

  • Performance Gains: MTSR-FLP consistently outperformed baselines. Compared to SASRec, the model improved HR@10 by 6.68% to 31.10% and NDCG@10 by 8.60% to 42.54%.
  • Domain Effectiveness: The most dramatic improvements occurred in the Steam dataset, suggesting the model is exceptionally effective in domains with strong temporal dependencies, such as gaming.
  • Ablation Insights: Ablation studies confirmed that the multi-temporal approach was the primary driver of success; models utilizing only a single type of temporal embedding showed significantly lower performance.

Conclusion

MTSR-FLP demonstrates that integrating multi-dimensional temporal context with fused local and global preferences significantly enhances sequential recommendation accuracy. Future work may involve incorporating Graph Neural Networks (GNNs) to capture higher-order relationships and expanding the feature set to include item categories and user demographics.

Summary updated

Abstract

Abstract Since sequential recommendation can effectively capture the user’s dynamic interaction pattern according to the user’s historical interaction, it plays one of the important roles in the field of recommendation systems. Nowadays, the sequential recommendation models based on deep learning approaches have achieved good results in reality due to their excellent performance in information retrieval and filtering. However, there are still some challenges that still should be overcome of them. A major disadvantage of them is that they are only considering users’ short-time interests while ignoring their long-term preferences. Moreover, they are incapable of considering the influence of time information in the original interaction sequence, which could be helpful to fully extract various patterns via the different temporal embedding forms. Therefore, this paper proposes a novel model named multi-temporal sequential recommendation model based on the fused learning preference named MTSR-FLP in short. In the proposed framework, this paper adopts an MLP-based state generation module to consider the user’s long-term preferences and the short-time interests simultaneously. In particular, the proposed MTSR-FLP model designs a global representation learning approach to obtain the user’s global preference, and a local representation learning approach to capture the users’ local preference via its historical information. Moreover, the proposed model develops a multiple temporal embedding scheme to encode the positions of user–item interactions of a sequence, in which multiple kernels are utilized for the absolute or relative timestamps to establish unique embedding matrices. Finally, compared with other advanced sequence recommendation models on five public real-world datasets, the experimental results show that the proposed MTSR-FLP model has improved the performance of HR@10 from the 6.68% through 31.10% and NDCG@10 from the 8.60% through 42.54%.
BibTeX
@article{chen2023multi,
  author = {Jianxia Chen and Liwei Pan and Shi Dong and Tianci Yu and Liang Xiao and Meihan Yao and Shijie Luo},
  title = {{Multi-temporal Sequential Recommendation Model Based on the Fused Learning Preferences}},
  journal = {International Journal of Computational Intelligence Systems},
  year = {2023},
  doi = {10.1007/s44196-023-00310-w},
}
Figure