Learning Hierarchical Representation Model for NextBasket Recommendation

By Pengfei Wang, Jiafeng Guo, Yanyan Lan, Jun Xu, Shengxian Wan, Xueqi Cheng. 2015

436 citations recorded by the source corpus.

Subfields: Information Systems, Marketing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Systems

1. Xiangnan He
2. Xiang Wang
3. Yongfeng Zhang
4. Wayne Xin Zhao
5. Tat-seng Chua
6. Xu Chen
7. Julian McAuley
8. Meng Wang

In Marketing

1. Yongfeng Zhang
2. Xu Chen
3. Xiangnan He
4. Zheng Qin
5. Tat-seng Chua
6. Hongyuan Zha
7. Hongteng Xu
8. Yixin Cao

Citations in the corpus

• SHINE: Signed Heterogeneous Information Network Embedding for Sentiment Link Prediction (2017, background): “eddingsintofinalheterogeneousembeddingei byspecificaggregation functionд(·,·,·). We list some of the available aggregation functions as follows: •Summation [34], i.e., ei =bxi +byi +bzi; •Max pooling [28], i.e., ei = element -wise max(bxi ,byi zi); •Concatenation [23], i.e., ei = ⟨bxi,byi,bzi⟩. Fina”
• Improving Sequential Recommendation with Knowledge-Enhanced Memory Networks (2018, background): “As two typical applications of representation learning, HRM [32] captures both sequential behavior and users’ general taste by involving transaction and user representations in prediction, and TransRec[10] proposes that items are embedded into a “transition space" where each user is modeled by a tra”
• CoNet: Collaborative Cross Networks for Cross-Domain Recommendation (2018, methodology): “Some hierarchical methods can reduce to factorize a specific matrix [40].”
• Neural Survival Recommender (2017, methodology): “Popular models on this line of research include FPMC [30] and HRM [32].”
• Dynamic Explainable Recommendation Based on Neural Attentive Models (2019, methodology): “The hierarchical representation model (HRM) (Wang et al. 2015) generalized FPMC into a representation learning framework, and significantly improved the recommendation performance.”
• Controllable Multi-Interest Framework for Recommendation (2020, methodology): “HRM [52] extends the FPMC model and employs a two-layer structure to construct a hybrid representation over users and items from the last transaction.”
• A Survey on Session-based Recommender Systems (2019, methodology): “, noisy sessions) [30],[43],[53],[109], [114],[117],[118]”
• STAMP: Short-Term Attention/Memory Priority Model for Session-based Recommendation (2018, background): “[19] proposed a hybrid representation learning model, which employs a two-layer hierarchical structure for modeling of the sequential behavior and general interests of users from their last transactions.”
• Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding (2018, background): “The Markov chain based model [2, 6, 21, 30] is an early approach to top-N sequential recommendation, where an L-order Markov chain makes recommendations based on L previous actions.”
• Neural Factorization Machines for Sparse Predictive Analytics (2017, background): “In contrast to traditional deep learning methods that simply concatenate [9, 31, 44] or average [16, 36] embedding vectors in the low level, our use of Bi-Interaction pooling encodes more informative feature interactions, greatly facilitating the following “deep” layers to learn meaningful…”
• Sequential Recommender Systems: Challenges, Progress and Prospects (2019, methodology): “Specifically, some works take the learned latent representations as the input of a network to further calculate an interaction score between users and items, or successive users’ actions [18; 19], while other works directly utilize them to calculate a metric like the Euclidean distance as the intera”
• Predicting the Next Location: A Recurrent Model with Spatial and Temporal Contexts (2016, background): “However, FPMC assumes that all the component are linearly combined, indicating that it makes strong independent assumption among factors (Wang et al. 2015).”
• Sequential Recommendation with User Memory Networks (2018, methodology): “Many works [23, 27] have shown that users’ recent behaviors can be more important to the current decisions.”
• A Dynamic Recurrent Model for Next Basket Recommendation (2016, background): “(5) HRM [9] is a state-of-the-art hierarchical representation model, which can capture general users’ interests and sequential effects.”
• Sequence-Aware Recommender Systems (2018, methodology): “he sliding window. Finally, they use a parallelized version of the Tree Augmented Naive Bayes algorithm for the classification (prediction) of the next used app. In the e-commerce domain, Wang et al. [110] use Feed-forward Neural Networks for next-basket recommendation. For the predictions, only the”
• TALLRec: An Effective and Efficient Tuning Framework to Align Large Language Model with Recommendation (2023, methodology): “Moving forward, we plan to explore more efficient methods to activate the recommendation ability of larger models and tune LLMs to handle multiple recommendation tasks simultaneously.”
• Time Interval Aware Self-Attention for Sequential Recommendation (2020, methodology): “PRME [6], HRM [27], Fossil [8], GRU4Rec[12]) have been outperformed by the baselines above, so we omit comparison against them.”