Multibehavior recommendation aims to leverage diverse user interaction types—such as views, clicks, and purchases—to better understand user intent and enhance recommendation accuracy. However, effectively modeling these heterogeneous behaviors remains challenging due to the sparsity, noise, over-smooth embedding, and dynamic nature of interaction sequences. To address these issues, we propose GCTMR (Graph Contrastive Transformer for Multibehavior Recommendation), a framework that integrates graph-based contrastive learning with a behavior-aware Transformer architecture. Specifically, the methodological contribution lies in three key innovations: 1) an adaptive sparse attention mechanism employing α -entmax that dynamically filters out noisy or irrelevant interactions to preserve the distinctiveness of user behavior representations; 2) a frequency-aware multibehavior encoder using FFT/IFFT decomposition that disentangles stable long-term preferences from short-term behavioral shifts across different interaction types; and 3) a simplified focal loss function that emphasizes hard-to-predict instances, improving robustness under data sparsity and behavioral ambiguity. The empirical validation is exceptionally strong, demonstrating improving of 22.43% on Tmall and 18.10% on Beibei over state-of-the-art baselines in HR@10 metrics, with particularly remarkable performance under sparse data conditions where traditional methods struggle most. Comprehensive ablation studies confirm that each component contributes meaningfully to the overall performance, while extensive hyperparameter analyses demonstrate the framework’s stability and practical applicability.

Social recommendation utilizes social relationships to alleviate data sparsity and improve recommendation quality. Nevertheless, existing approaches still encounter several limitations. First, user preference structures in social networks are often hierarchical and non-Euclidean, whereas most existing models rely on Euclidean embeddings that are insufficient for capturing such complex geometries. Second, noisy social connections may propagate unreliable information and degrade representation quality. Third, commonly used multi-view fusion strategies based on multilayer perceptrons (MLPs) often lack the expressive capability required to effectively integrate heterogeneous user signals. To address these challenges, this paper proposes HGKAN (Hyperbolic Graph-Based Kolmogorov–Arnold Network), a unified framework that combines hyperbolic representation learning, graph neural networks, and KAN-based nonlinear fusion for social recommendation. Specifically, a Hyperbolic Intent Attention module is introduced to learn fine-grained and hierarchical user intent representations in hyperbolic space. In addition, a Hyperbolic Attention GCN is designed to denoise and aggregate social information while preserving complex relational structures. Furthermore, a KAN-based Multi-View Fusion Layer is developed to replace conventional MLP-based fusion, enabling more expressive and interpretable integration of preference, social, and intent representations. Extensive experiments conducted on three public benchmark datasets, including Yelp, Ciao, and Douban, demonstrate that HGKAN consistently outperforms several state-of-the-art recommendation methods. Experimental results show that the proposed framework achieves substantial improvements in both Recall@5 and NDCG@5, with all performance gains being statistically significant (p < 0.05). These findings verify the effectiveness of integrating hyperbolic geometric learning with KAN-based multi-view fusion for robust and accurate social recommendation.

Sequential behaviors and social ties jointly shape user preferences; however, most prior work models them in isolation and relies on shallow fusion in Euclidean space, which struggles to capture temporal drift and hierarchical social structure. We propose a novel framework; HGT4Rec, a Hyperbolic Graph Transformer for Sequential and Social Recommendation. A graph transformer encodes item-transition dependencies to track evolving preferences along the sequence, while a hyperbolic transformer operates on the social graph to represent long-range and hierarchical influence. We further introduce FusionGRU, an adaptive gating module that integrates the two representations into a unified preference state for Top-K prediction. Experiments on Yelp, iFashion, LastFM show that HGT4Rec delivers substantial improvements over the strongest baseline, achieving +348.22% / +314.86%, +383.21% / +210.19%, and +90.85% / +9.83% in Recall@10 and NDCG@10, respectively. Our results demonstrate the value of combining graph transformers with hyperbolic space modeling and the gated fusion for next-item recommendation in sequential and social settings (our code).

Sequential-social recommendation systems are essential for understanding users’ evolving interests and predicting their future behaviors. While existing methods employing bidirectional graph modeling have shown promising results in capturing user interactions, they face significant challenges in effectively leveraging bidirectional information and addressing dynamic user preferences. To overcome these limitations, we propose Bidirectional Transformer Graphs for Sequential-Social Recommendation (BiTG4Rec), a novel framework that integrates both supervised and self-supervised learning to analyze dynamic user preferences at sequential and social levels through bidirectional graphs. Our approach comprises three core components: 1) a Bidirectional Dynamic Graph Convolutional Network (BiDGCN) that models long-term sequential preferences, 2) a Bidirectional Dynamic Graph Attention Network (BiDGAT) that captures short-term sequential preferences, and 3) a Bidirectional Hyperbolic Graph Contrastive Learning module (BiHGCL) that extracts social preferences. To unify these diverse signals, we introduce a TransformerGated mechanism that dynamically integrates long-term, short-term, and social preferences. Extensive experiments on multiple real-world datasets demonstrate that BiTG4Rec consistently outperforms state-of-the-art methods, validating its effectiveness for sequential-social recommendation tasks.

The integration of sequential and social recommendations using a graph-based co-attention architecture marks a significant advancement in deep learning, substantially improving recommender system performance. This innovation enables effective representation learning of user interactions by capturing dynamic preferences influenced by both sequential and social contexts. While prior research has attempted to model evolving user preferences, existing approaches suffer from two key limitations: 1) ineffective modeling of complex sequential dependencies and 2) overlooking the hierarchical nature of social influence, resulting in suboptimal performance. To address these challenges, we propose GCA4Rec (Graph-Based Co-Attention Networks for Sequential and Social Recommendation), a novel framework designed to handle the complexities of user preference dynamics across sequential and social domains. Our model combines a graph attention contrastive learning module to track sequential preference shifts, while a hyperbolic graph attention isomorphism network models social-level preference dynamics. Additionally, we introduce fusionGated, a novel gating mechanism that effectively integrates co-attention signals from both levels. Extensive experiments on real-world benchmark datasets demonstrate that GCA4Rec outperforms state-of-the-art methods in Top-k recommendation tasks and exhibits robustness under varying degrees of data sparsity.

This research presents an integrated machine learning approach for optimizing digital workforce matching in Thailand's evolving digital economy. The study develops a novel job recommendation system combining Natural Language Processing (NLP) with Random Forest classification to analyze job market data from Thailand's leading recruitment platforms. Using FastText for initial job classification and a Random Forest model for skill-based matching, the system achieves 75% accuracy in job recommendations across 20 digital job categories. The methodology incorporates automated skill extraction, cross-validated model comparison, and a user-friendly web interface for practical applications. Our findings reveal distinct skill clusters and job-skill relationships in Thailand's digital sector, with the Random Forest model outperforming traditional Decision Tree approaches by 4% in accuracy metrics. The system demonstrates robust performance in real-world testing, achieving 86.67% accuracy in matching previously unseen job postings. This research contributes to both theoretical understanding of skill-based job matching and practical workforce development, offering insights for curriculum development and career planning for workforce development stakeholders in Thailand's digital sector.