Large Language Models (LLMs) offer groundbreaking advancements in recommender systems through superior text analysis and decision-making support. However, integrating LLMs into recommender systems still suffers from the problems of identifier uninterpretability and lack of transparency. To address these issues and fully leverage the capabilities of LLMs, we propose a chain of thought (CoT) based recommendation framework called CoT4Rec which employs LLMs as data enhancers for user preference analysis. Initially, we design a CoT reasoning strategy that can derive more behaviorally-aligned user preference features by clustering users’ historical interactions. Subsequently, we propose a two-stage recommendation model that not only makes full use of the world knowledge embedded in LLMs but also generates a logically transparent reasoning path. By integrating a user preference analyzer early in the recommendation pipeline, the model deeply analyzes users' historical interactions, helping to enhance the personalization and transparency of the recommender system. CoT4Rec demonstrates superior performance over existing state-of-the-art models in recommendation tasks across four public datasets, achieving improvements ranging from 2.2% to 12.2%.

Next Point-of-Interest (POI) recommendation, which predicts users' future destinations based on their potential interests, has emerged as a critical task in location-based social networks (LBSNs). However, this task remains challenged by issues such as popularity bias, exposure bias, and limited representational capacity, all of which impede the accurate modeling of users and POIs, thereby restricting balanced and effective recommendations. Therefore, we propose Trajectory Collaborative Filtering Mamba (TCFMamba), which integrates two specially designed modules, i.e., Joint Learning of Static and Dynamic Representations (JLSDR) and Preference State Mamba Network (PSMN), for debiased Point-of-Interest recommendation.

Extreme precipitation, as a core causative factor of meteorological disasters, poses significant challenges for accurate short-term forecasting due to the chaotic nature of precipitation systems and their multi-scale spatio-temporal evolution. Traditional numerical models are notably affected by error accumulation, while existing deep learning models still face dual limitations in physical fidelity and multi-scale feature extraction. To Address these issues, we propose an innovative Multi-scale Physics-informed Transformer with spatio-temporal feature adapter for extreme precipitation nowcasting, termed MPFormer. Our framework comprises two core components: the deterministic Evolution Network and the stochastic Generative Network. The Evolution Network integrates a novel Scale-Aware Temporal Residual Modulation Transformer (STRMT) encoder that captures multi-scale storm dynamics through residual temporal attention. The Generative Network introduces spatio-temporal adapters as lightweight transfer modules for probabilistic modeling. We develop a Multi-scale Physics-informed Loss with three innovations: 1) dynamic weight scheduling for feature fusion, 2) physical constraints preserving storm evolution patterns, and 3) entropy-based uncertainty calibration. Experiments based on MRMS radar data from North America demonstrate that the model can generate high-resolution forecasts (2km grid) with a 3-hour lead time over an area of 2048×2048 square kilometers. Compared to the state-of-the-art technologies, the proposed framework shows significant effectiveness and superiority in metrics such as CSIN, offering a new paradigm that combines physical interpretability with engineering practicality for extreme weather warnings and disaster prevention in smart cities.

The advent of single-cell RNA-sequencing (scRNA-seq) technology promotes biological analysis at the cellular level. Clustering cells to identify the type of cell is an important step in scRNA-seq analysis. Most of the existing clustering methods based on deep learning technology first adopt an autoencoder-decoder module to learn the low-dimensional features of cells and then apply other modules to learn the clustering relationship features of cells. However, the two-stage learning process makes the model training more difficult. Here we propose a novel cell representation learning method that is based on a local self-attention network and contrastive learning for scRNA-seq clustering. In particular, a local self-attention network automatically aggregates potential information of cells based on a cell relationship graph, and a dual contrastive learning module simultaneously optimizes the cell representation in cell- and cluster-level. The cell-level module makes related cells similar at the feature level, whereas the cluster-level module enables cells to form clusters at the cluster level. Finally, the powerful Leiden community discovery algorithm is used for clustering based on learned representation. In brief, we construct cell pairs through cell relationships and utilize contrastive learning to directly learn cell representations in a low-dimensional space while preserving their local structural relationships without pretraining an autoencoder-decoder module. Three benchmark experiments on 160 subsample datasets with different numbers of cell types, 3 datasets of different protocols, and 9 real public datasets demonstrate the superior performance of the proposed method compared with baseline methods.

Trajectory inference is a crucial task in single-cell RNA-sequencing downstream analysis, which can reveal the dynamic processes of biological development, including cell differentiation. Dimensionality reduction is an important step in the trajectory inference process. However, most existing trajectory methods rely on cell features derived from traditional dimensionality reduction methods, such as principal component analysis and uniform manifold approximation and projection. These methods are not specifically designed for trajectory inference and fail to fully leverage prior information from upstream analysis, limiting their performance. Here, we introduce scCRT, a novel dimensionality reduction model for trajectory inference. In order to utilize prior information to learn accurate cells representation, scCRT integrates two feature learning components: a cell-level pairwise module and a cluster-level contrastive module. The cell-level module focuses on learning accurate cell representations in a reduced-dimensionality space while maintaining the cell-cell positional relationships in the original space. The cluster-level contrastive module uses prior cell state information to aggregate similar cells, preventing excessive dispersion in the low-dimensional space. Experimental findings from 54 real and 81 synthetic datasets, totaling 135 datasets, highlighted the superior performance of scCRT compared with commonly used trajectory inference methods. Additionally, an ablation study revealed that both cell-level and cluster-level modules enhance the model's ability to learn accurate cell features, facilitating cell lineage inference. The source code of scCRT is available at https://github.com/yuchen21-web/scCRT-for-scRNA-seq.

Traditional music recommender systems are mainly based on users’ interactions, which limit their performance. Particularly, various kinds of content information, such as metadata and description can be used to improve music recommendation. However, it remains to be addressed how to fully incorporate the rich auxiliary/side information and effectively deal with heterogeneity in it. In this paper, we propose a M ulti-view E nhanced G raph A ttention N etwork (named MEGAN ) for session-based music recommendation. MEGAN can learn informative representations (embeddings) of music pieces and users from heterogeneous information based on graph neural network and attention mechanism. Specifically, the proposed approach MEGAN firstly models users’ listening behaviors and the textual content of music pieces with a Heterogeneous Music Graph (HMG). Then, a devised Graph Attention Network is used to learn the low-dimensional embedding of music pieces and users and by integrating various kinds of information, which is enhanced by multi-view from HMG in an adaptive and unified way. Finally, users’ hybrid preferences are learned from users’ listening behaviors and music pieces that satisfy users real-time requirements are recommended. Comprehensive experiments are conducted on two real-world datasets, and the results show that MEGAN achieves better performance than baselines, including several state-of-the-art recommendation methods.

Recommender systems are important approaches for dealing with the information overload problem in the big data era, and various kinds of auxiliary information, including time and sequential information, can help improve the performance of retrieval and recommendation tasks. However, it is still a challenging problem how to fully exploit such information to achieve high-quality recommendation results and improve users' experience. In this work, we present a novel sequential recommendation model, called multivariate Hawkes process embedding with attention (MHPE-a), which combines a temporal point process with the attention mechanism to predict the items that the target user may interact with according to her/his historical records. Specifically, the proposed approach MHPE-a can model users' sequential patterns in their temporal interaction sequences accurately with a multivariate Hawkes process. Then, we perform an accurate sequential recommendation to satisfy target users' real-time requirements based on their preferences obtained with MHPE-a from their historical records. Especially, an attention mechanism is used to leverage users' long/short-term preferences adaptively to achieve an accurate sequential recommendation. Extensive experiments are conducted on two real-world datasets (lastfm and gowalla), and the results show that MHPE-a achieves better performance than state-of-the-art baselines.

Recommender systems, which aim to provide personalized suggestions for users, have proven to be an effective approach to cope with the information overload problem existing in many online applications and services. In this paper, we target two specific sequential recommendation tasks,next music recommendation and next new music recommendation, to predict the next (new) music piece that users would like based on their historical listening records. In current music recommender systems, various kinds of auxiliary/side information, e.g., item contents and users’ contexts, have been taken into account to facilitate user/item preference modeling and have yielded comparable performance improvement. Despite the gained benefits, it is still a challenging and important problem to fully exploit sequential music listening records due to the complexity and diversity of interactions and temporal contexts among users and music, as well as the dynamics of users’ preferences. To this end, this paper proposes a novelAttentiveTemporalPointProcess (ATPP) approach for sequential music recommendation, which is mainly composed of a temporal point process model and an attention mechanism. OurATPPcan effectively capture the long- and short-term preferences from the sequential behaviors of users for sequential music recommendation. Specifically,ATPPis able to discover the complex sequential patterns from the interaction between users and music with the temporal point process, as well as model the dynamic impact of historical music listening records on next (new) music pieces adaptively with an attention mechanism. Comprehensive experiments on four real-world music datasets demonstrate that the proposed approachATPPoutperforms state-of-the-art baselines in both next and next new music recommendation tasks.

Traditional recommendation methods suffer from limited performance, which can be addressed by incorporating abundant auxiliary/side information. This article focuses on a personalized music recommender system that incorporates rich content and context data in a unified and adaptive way to address the abovementioned problems. The content information includes music textual content, such as metadata, tags, and lyrics, and the context data incorporate users' behaviors, including music listening records, music playing sequences, and sessions. Specifically, a heterogeneous information network (HIN) is first presented to incorporate different kinds of content and context data. Then, a novel method called content- and context-aware music embedding (CAME) is proposed to obtain the low-dimension dense real-valued feature representations (embeddings) of music pieces from HIN. Especially, one music piece generally highlights different aspects when interacting with various neighbors, and it should have different representations separately. CAME seamlessly combines deep learning techniques, including convolutional neural networks and attention mechanisms, with the embedding model to capture the intrinsic features of music pieces as well as their dynamic relevance and interactions adaptively. Finally, we further infer users' general musical preferences as well as their contextual preferences for music and propose a content- and context-aware music recommendation method. Comprehensive experiments as well as quantitative and qualitative evaluations have been performed on real-world music data sets, and the results show that the proposed recommendation approach outperforms state-of-the-art baselines and is able to handle sparse data effectively.

We introduce a new approach for performing image inpainting, by devising an integrative method based on the superpixel segmentation technique and considering minimal user input. Image inpainting methods are concerned with filling in missing or replacing undesired regions in an image. Typically, inpainting methods consider and extrapolate known image data. Superpixels in the immediate neighborhood of the in-painting region are computed and used as source image data to fill in (or replace) the inpainting area. A user provides additional information by specifying line segments in the image to assist the otherwise automatic inpainting process, to ensure that only desirable superpixels are utilized when copying them into the inpainting region. User interaction is minimal, as it is merely necessary to specify a small number of line segments that define image parts to be used as source data in distinct inpainting regions. We provide experimental results demonstrating that our method performs well when compared against other methods, especially concerning the preservation of edges and texture in the inpainted regions.

Example-based super-resolution algorithms, which predict unknown high-resolution image information using a relationship model learnt from known high- and low-resolution image pairs, have attracted considerable interest in the field of image processing. In this paper, we propose a multi-example feature-constrained back-projection method for image super-resolution. Firstly, we take advantage of a feature-constrained polynomial interpolation method to enlarge the low-resolution image. Next, we consider low-frequency images of different resolutions to provide an example pair. Then, we use adaptive k NN search to find similar patches in the low-resolution image for every image patch in the high-resolution low-frequency image, leading to a regression model between similar patches to be learnt. The learnt model is applied to the low-resolution high-frequency image to produce high-resolution high-frequency information. An iterative back-projection algorithm is used as the final step to determine the final high-resolution image. Experimental results demonstrate that our method improves the visual quality of the high-resolution image.

Contextual factors can benefit music recommendation and retrieval tasks remarkably. However, how to acquire and utilize the contextual information still need to be studied. In this paper, we propose a context aware music recommendation approach, which can recommend music appropriate for users' contextual preference for music. In analogy to matrix factorization methods for collaborative filtering, the proposed approach does not require songs to be described by features beforehand, but it learns music pieces' embeddings (vectors in low-dimensional continuous space) from music playing records and corresponding metadata and infer users' general and contextual preference for music from their playing records with the learned embedding. Then, our approach can recommend appropriate music pieces. Experimental evaluations on a real world dataset show that the proposed approach outperforms baseline methods.

Image super‐resolution (SR) for a single low‐resolution image is an important and challenging task in image processing. In this study, the authors propose a novel non‐local feature back‐projection method for image SR, which can effectively reduce jaggy and ringing artefacts common, in general, iterative back‐projection (IBP) method. In their method, the objective high‐resolution (HR) image is obtained by projecting reconstructed errors back to HR image iteratively. To optimise the initial HR image and constrain anisotropic errors propagation during IBP process, an efficient non‐local feature interpolation algorithm is designed. Specially, edge information is used as constraints to make the interpolation surface preserve better shape. Furthermore, as post‐processing, non‐local similarities are utilised to remove noise and irregularities induced by errors propagation. Experimental results show that their method achieves better performance than state‐of‐the‐art methods in terms of both quantitative metrics and visual qualities.

Conventional polynomial interpolation methods produce images with blurred edges, while edge-directed interpolation methods make enlarged images with good quality edges but with detail distortion in the non-edge portion. A new method for constructing a fitting surface to image data is presented. Unlike existing methods which produce enlarged images using image data as interpolation data, the new method constructs the fitting surface using the image data as constraints to reverse the sampling process for improving the fitting precision. To remove the zigzagging artifact, for each pixel and its nearby region, the edge information is used to determine the quadratic polynomial which approximates the original scene with a quadratic polynomial precision. Comparison results of the new method with other methods are included.