Multimodal Sequential recommendation alleviates the semantic insufficiency and data sparsity of item-ID-based models by incorporating side information such as text and images. However, multimodal systems face the dual challenges of feature-space heterogeneity and modality-specific noise, in addition to the dynamic evolution of user interests over time. Existing methods still struggle to jointly handle cross-modal alignment and time-aware preference modeling. To address these challenges, we propose a multimodal sequential recommendation framework with cross-modal alignment and temporal gating, which leverages item ID, text, and image modalities to capture users’ dynamic interests. The proposed model contains three core components. First, a cross-modal alignment mixture-of-experts module preserves modality-specific features with dedicated experts and captures shared semantics with common experts, thereby mitigating the semantic mismatch inherent in direct fusion. Second, a hierarchical time-aware mixture-of-experts module uses short-term intervals, long-term spans, and periodic time encodings for expert routing, and applies a time-aware modality gate to adaptively adjust the importance of ID, text, and image modalities under different temporal contexts. Third, a sequential interest contrastive learning objective enhances the discriminability of ID-based sequential interest representations by leveraging dynamic temperature scaling, multi-scale positive samples, hard negative mining, and diversity regularization. Experiments on games, beauty, and toys demonstrate that the proposed method consistently outperforms representative sequential and multimodal recommendation baselines on Normalized Discounted Cumulative Gain (NDCG)@5, NDCG@10, Mean Reciprocal Rank (MRR)@5, and MRR@10. Furthermore, ablation results validate the effectiveness of each proposed component.
Nowadays, the volume of heterogeneous data in situational analysis centers is growing exponentially, leading to information overload for decision makers (DMs) and a decrease in the effectiveness of traditional decision support systems (DSS). Intelligent DSSs (ISDSS) demonstrate potential, but face challenges in explainability, heterogeneous data integration, cognitive load, and scalability. This paper proposes a method for intelligent decision support focused on identifying and generating options for resolving emergency situations—conditions that have no exact precedents in the knowledge base. The method includes formalizing the situation using a vector of normalized parameters St, separating it into independent and dependent variables with the construction of a dependency tree, neural network classification of three types of conditions (normal, abnormal, and emergency) with a forecast for a lead interval τ, the synthesis of solutions for emergency situations based on an analysis of proximity graphs to known emergency precedents and evolutionary optimization. A computational experiment was conducted on the open dataset of the Tennessee Eastman Process simulation model with 28 failure types and 200 repeated simulations. The neural network classifier achieved an accuracy of 0.88 and a macro-averaged F1-score of 0.87 on a test set of 200 situations. Graphs of nearby emergency precedents were constructed for 50 synthetic emergency situations; analysis demonstrated the stability of topological characteristics (vertex degree 5.62 ± 1.18, closeness centrality 0.43 ± 0.09), substantiating the applicability of graph neural networks for accelerated control action synthesis. The proposed method reduces dependence on expert assessments and improves the adaptability and explainability of decisions, while the demonstrated stability of the graph-based precedent retrieval lays the groundwork for future full-scale validation of control-action synthesis in next-generation hybrid IDSS.
Against the backdrop of digital transformation in higher education, campus information is often dispersed across independent platforms, creating fragmented access and retrieval barriers. To address these challenges, this study proposes a multimodal agentic RAG framework for integrating multi-source, heterogeneous university information through a unified natural language interaction portal. Compared with basic RAG architectures, the framework introduces three improvements: (1) vision–language models convert unstructured visual content into indexable textual evidence; (2) BM25 sparse retrieval, dense-vector retrieval, and real-time online retrieval jointly support keyword matching, semantic retrieval, and up-to-date information access; and (3) agent-based dynamic routing adaptively schedules retrieval tools according to query intent and evidence sufficiency, reducing redundant calls. Experiments show that the proposed framework achieves higher overall performance in answer accuracy, factual consistency, evidence coverage, and composite score than general-purpose large language models, traditional single-path RAG, and representative advanced RAG methods including Self-RAG, Adaptive-RAG, and VisRAG. Ablation studies further validate the contributions of the core modules. Overall, the framework provides a low-disruption solution for unified information retrieval and natural language interaction in higher education scenarios.
Multi-view representation learning, aiming to extract discriminative patterns from heterogeneous data sources, has attracted increasing attention in recent years. A central challenge in multi-view data fusion lies in determining the appropriate weight and importance of each view in a principled manner. While graph-based methodologies have proven effective for multi-view learning, the systematic development of optimization-centric methodologies for consistent graph fusion remains in its nascent stages. Furthermore, the lack of explicit alignment between the latent feature representation space and the graph space is frequently overlooked, ultimately compromising the consistency of the shared representation. To this end, we propose a bi-level framework for joint multi-view feature representation learning and graph fusion, which formally formulates the multi-view data fusion task as a rigorous mathematical optimization problem. The proposed approach demonstrates consistent effectiveness in clustering tasks, achieving state-of-the-art performance on six benchmark datasets with accuracy improvements ranging up to 9.73% over existing methods.
This paper studies whether structured representations can retain the meaning contained in scientific sentences. A lightweight LLM is fine-tuned to generate JSON structures from scientific sentences. The objective was to capture information contained in the sentence using a hierarchical structure. The structured representations are then used to reconstruct sentences with a generative model to test whether such representations are useful to retain the meaning of the sentence. The evaluation compares the reconstructed sentences with the originals using semantic and lexical similarity. Our results show that hierarchical structured formats preserve semantic information with an average cosine similarity of 0.87.
