Shared-account Cross-domain Sequential Recommendation (SCSR) is an emerging yet challenging task that simultaneously considers the shared-account and cross-domain characteristics in sequential recommendation. Existing works on SCSR are mainly based Recurrent Neural Network (RNN) Graph (GNN) but they ignore fact although multiple users share a single account, it occupied by one user at time. This observation motivates us to learn more accurate user-specific account representation attentively...

Shared-account cross-domain sequential recommendation (SCSR) task aims to recommend the next item via leveraging mixed user behaviors in multiple domains. It is gaining immense research attention as more and users tend sign up on different platforms share accounts with others access domain-specific services. Existing works SCSR mainly rely mining patterns recurrent neural network (RNN)-based models, which suffer from following limitations: 1) RNN-based methods overwhelmingly target...

Since Facebook announced itself as Meta, the metaverse has been popular with educationalists. However, scanty studies have bibliographically analyzed use of in education. This study complements missing link literature by bibliometrically analyzing research into education using both VOSviewer and CitNetExplorer. The identified top authors, organizations, countries, keywords, documents, sources, trends, challenges used education, together effects on learning environments, interactions,...

Graph-based Sequential Recommender systems (GSRs) have gained significant research attention due to their ability simultaneously handle user-item interactions and sequential relationships between items. Current GSRs often utilize composite or in-depth structures for graph encoding (e.g., the Graph Transformer). Nevertheless, they high computational complexity, hindering deployment on resource-constrained edge devices. Moreover, relative position in Transformer has difficulty considering...

Shared-account Sequential Recommendation (SSR) aims to provide personalized recommendations for accounts shared by multiple users with varying sequential preferences. Previous studies on SSR struggle capture the fine-grained associations between interactions and different latent within account's hybrid sequences. Moreover, most existing methods (e.g., RNN-based or GCN-based methods) have quadratic computational complexities, hindering deployment of SSRs resource-constrained devices. To this...

Load prediction tasks aim to predict the dynamic trend of future load based on historical performance sequences, which are crucial for cloud platforms make timely and reasonable task scheduling. However, existing models limited while capturing complicated temporal patterns from sequences. Besides, frequently adopted global weighting strategy (e.g., self-attention mechanism) in modeling schemes has quadratic computational complexity, hindering immediate response servers complex real-time...

Shared-account Sequential Recommendation (SSR) aims to provide personalized recommendations for accounts shared by multiple users with varying sequential preferences. Previous studies on SSR struggle capture the fine-grained associations between interactions and different latent within account's hybrid sequences. Moreover, most existing methods (e.g., RNN-based or GCN-based methods) have quadratic computational complexities, hindering deployment of SSRs resource-constrained devices. To this...

Cross-domain sequential recommenders (CSRs) are gaining considerable research attention as they can capture user preference by leveraging side information from multiple domains. However, these works typically follow an ideal setup, i.e., different domains obey similar data distribution, which ignores the bias brought asymmetric interaction densities (a.k.a. inter-domain density bias). Besides, frequently adopted mechanism (e.g., self-attention network) in sequence encoder only focuses on...

Preventive protection of cultural relics is to make use all the science and technology beneficial research archaeological heritage predict disease relics. The existing preventive system has made some achievements in environmental monitoring, but analysis utilization large data are still insufficient. In this paper, under idea multisource information fusion, a least squares support vector machine regression method based on multivariate time series wavelet correlation proposed achieve accurate...

Cross-domain Sequential Recommendation (CSR) is an emerging yet challenging task that depicts the evolution of behavior patterns for overlapped users by modeling their interactions from multiple domains. Existing studies on CSR mainly focus using composite or in-depth structures achieve significant improvement in accuracy but bring a huge burden to model training. Moreover, learn user-specific sequence representations, existing works usually adopt global relevance weighting strategy (e.g.,...

Cross-domain sequential recommenders (CSRs) are gaining considerable research attention as they can capture user preference by leveraging side information from multiple domains. However, these works typically follow an ideal setup, i.e., different domains obey similar data distribution, which ignores the bias brought asymmetric interaction densities (a.k.a. inter-domain density bias). Besides, frequently adopted mechanism (e.g., self-attention network) in sequence encoder only focuses on...

Due to an exponential increase in the number of academic literature resources, difficulty resource integration and utilization becomes even greater. Especially, aggregation resources different sources, formats at diverse granularities is now confronted with new challenges. At present, industrial users have put forward a requirement rapidly acquiring high quality information resources. To meet this requirement, we explored vertical domain oriented model. By virtue model, it expected solve...