As global e-commerce platforms continue to expand, companies are entering new markets where they encounter cold-start challenges due to limited human labels and user behaviors. In this paper, we share our experiences in Coupang to provide a competitive cold-start performance of relevance matching for emerging e-commerce markets. Specifically, we present a Cold-Start Relevance Matching (CSRM) framework, utilizing a multilingual Large Language Model (LLM) to address three challenges: (1) activating cross-lingual transfer learning abilities of LLMs through machine translation tasks; (2) enhancing query understanding and incorporating e-commerce knowledge by retrieval-based query augmentation; (3) mitigating the impact of training label errors through a multi-round self-distillation training strategy. Our experiments demonstrate the effectiveness of CSRM-LLM and the proposed techniques, resulting in successful real-world deployment and significant online gains, with a 45.8% reduction in defect ratio and a 0.866% uplift in session purchase rate.

Identifying the miRNAs that are associated with diseases can assist to explore the pathogenesis of diseases. Traditional prediction methods primarily focus on integrating multi-sourced data related to miRNAs and diseases within euclidean space to infer potential candidate disease-related miRNAs. Research indicates that miRNAs belonging to the same family, residing in the same cluster, or sharing more common target proteins are more likely to be involved into similar disease processes. However, existing approaches have not fully integrated the family, cluster, and common target protein attributes of miRNAs, nor explored the low-frequency smoothness and high-frequency local fluctuation features of miRNA and disease nodes in the frequency domain. To overcome these issues, we propose a Fourier transform and Kolmogorov-Arnold network enhanced relation-aware generative adversarial network (FKRGAN) model. FKRGAN incorporates a Fourier transform enhanced dual-space feature learning (FDFL) strategy, which helps learn the topological features of miRNA and disease nodes in euclidean space, as well as their low-frequency and high-frequency characteristics in the frequency domain. Furthermore, a feature-level attention mechanism is designed to determine the significance of features learned from dual-space representations and the family, cluster, and target protein features derived from homogeneous graphs, thereby facilitating the adaptive fusion of these features. We develop a relation-aware generative adversarial network with Kolmogorov-Arnold networks (KRGAN), which enhances feature learning for each miRNA and disease node by employing generative and adversarial strategies. The generator, composed of multi-layer Kolmogorov-Arnold networks (KAN), fuses multiple connection relationships between miRNA and disease nodes during the generation process, effectively capturing nonlinear dependencies among node features to produce relation-aware node feature representations. Comparative experiments on public datasets demonstrate that our method outperforms eight state-of-the-art prediction methods. Case studies on three diseases further show FKRGAN's ability to identify candidate miRNA-disease associations.

Matching preferred shows to the subscribers is extremely important in the Over-the-Top (OTT) platforms. The existing methods did not adequately consider the characteristics of the OTT services, i.e., rich meta information, diverse user interests, and mixed recommendation scenarios, leading to sub-optimal performance. This paper introduces the Multi-Modal Multi-Interest Multi-Scenario Matching (M5) for the OTT recommendation to fully exploit these attributes. A multi-modal embedding layer is first introduced to transform the show IDs into both ID embeddings initialized randomly and content graph (CG) embeddings derived from the node representations pre-trained on a metagraph. To segregate the semantics between ID and CG embeddings, M5 exploits the mirrored two-tower modeling in the subsequent layers for efficiency and effectiveness. Specifically, a multi-interest extraction layer is proposed separately on ID and CG behaviors to model users' coarse-grained and fine-grained interests through behavioral categorization, subsidiary decoration, masked-language-modeling augmented self-attention modeling and subsidiary-intensity interest calibration. Facing the inherent diverse scenarios, M5 distinguishes the scenario differences at both feature and model levels, which crosses features with the scenario indicators and employs Split Mixture-of-Experts to generate the ID, and CG user embeddings. Finally, a weighted candidate matching layer is established to calculate the ID- and CG-oriented user-item preferences and then merge into a hybrid score with dynamic weighting. The extensive online and offline experiments over two real-world OTT platforms Hulu and Disney+ reveal that M5 significantly outperforms the previous state-of-the-art and online matching algorithms over various scenarios, indicating the effectiveness and robustness of the proposed method. M5 has been fully deployed on the main traffic of the most popular "For You'' sets of both platforms, continuously enhancing the user experience for hundreds of millions of subscribers every day and steadily increasing business revenue.

Content cold-start is a core problem in recommendation field, by which service providers can mine the potential profit from content that has not yet been discovered by most users, and provide more accurate personalized service to their users. In video recommendation, video and audio features should cover enough semantic information in the purpose of recommendation, thus should take an non-negligible role for content cold-start. This paper summarizes the Content Based Video Relevance Prediction Challenge held by Hulu, a top online streaming video platform in US, in ACM Multimedia conference 2019. The challenge is a content-based CTR prediction task for video recommendation, where millions of user interaction data and thousands of video features are released for research purpose on related topics.

Intrinsic Delaunay triangulation (IDT) naturally generalizes Delaunay triangulation from R 2 to curved surfaces. Due to many favorable properties, the IDT whose vertex set includes all mesh vertices is of particular interest in polygonal mesh processing. To date, the only way for constructing such IDT is the edge-flipping algorithm, which iteratively flips non-Delaunay edges to become locally Delaunay. Although this algorithm is conceptually simple and guarantees to terminate in finite steps, it has no known time complexity and may also produce triangulations containing faces with only two edges. This article develops a new method to obtain proper IDTs on manifold triangle meshes. We first compute a geodesic Voronoi diagram (GVD) by taking all mesh vertices as generators and then find its dual graph. The sufficient condition for the dual graph to be a proper triangulation is that all Voronoi cells satisfy the so-called closed ball property. To guarantee the closed ball property everywhere, a certain sampling criterion is required. For Voronoi cells that violate the closed ball property, we fix them by computing topologically safe regions, in which auxiliary sites can be added without changing the topology of the Voronoi diagram beyond them. Given a mesh with n vertices, we prove that by adding at most O ( n ) auxiliary sites, the computed GVD satisfies the closed ball property, and hence its dual graph is a proper IDT. Our method has a theoretical worst-case time complexity O ( n 2 + tn log n ), where t is the number of obtuse angles in the mesh. Computational results show that it empirically runs in linear time on real-world models.

Computing centroidal Voronoi tessellations (CVT) has many applications in computer graphics. The existing methods, such as the Lloyd algorithm and the quasi-Newton solver, are efficient and easy to implement; however, they compute only the local optimal solutions due to the highly non-linear nature of the CVT energy. This paper presents a novel method, called manifold differential evolution (MDE), for computing globally optimal geodesic CVT energy on triangle meshes. Formulating the mutation operator using discrete geodesics, MDE naturally extends the powerful differential evolution framework from Euclidean spaces to manifold domains. Under mild assumptions, we show that MDE has a provable probabilistic convergence to the global optimum. Experiments on a wide range of 3D models show that MDE consistently out-performs the existing methods by producing results with lower energy. Thanks to its intrinsic and global nature, MDE is insensitive to initialization and mesh tessellation. Moreover, it is able to handle multiply-connected Voronoi cells, which are challenging to the existing geodesic CVT methods.

Delaunay meshes (DM) are a special type of triangle mesh where the local Delaunay condition holds everywhere. We present an efficient algorithm to convert an arbitrary manifold triangle mesh M into a Delaunay mesh. We show that the constructed DM has O ( Kn ) vertices, where n is the number of vertices in M and K is a model-dependent constant. We also develop a novel algorithm to simplify Delaunay meshes, allowing a smooth choice of detail levels. Our methods are conceptually simple, theoretically sound and easy to implement. The DM construction algorithm also scales well due to its O ( nK log K ) time complexity. Delaunay meshes have many favorable geometric and numerical properties. For example, a DM has exactly the same geometry as the input mesh, and it can be encoded by any mesh data structure. Moreover, the empty geodesic circumcircle property implies that the commonly used cotangent Laplace-Beltrami operator has non-negative weights. Therefore, the existing digital geometry processing algorithms can benefit the numerical stability of DM without changing any codes. We observe that DMs can improve the accuracy of the heat method for computing geodesic distances. Also, popular parameterization techniques, such as discrete harmonic mapping, produce more stable results on the DMs than on the input meshes.

Computing geodesic distances on triangle meshes is a fundamental problem in computational geometry and computer graphics. To date, two notable classes of algorithms, the Mitchell-Mount-Papadimitriou (MMP) algorithm and the Chen-Han (CH) algorithm, have been proposed. Although these algorithms can compute exact geodesic distances if numerical computation is exact, they are computationally expensive, which diminishes their usefulness for large-scale models and/or time-critical applications. In this paper, we propose the fast wavefront propagation (FWP) framework for improving the performance of both the MMP and CH algorithms. Unlike the original algorithms that propagate only a single window (a data structure locally encodes geodesic information) at each iteration, our method organizes windows with a bucket data structure so that it can process a large number of windows simultaneously without compromising wavefront quality. Thanks to its macro nature, the FWP method is less sensitive to mesh triangulation than the MMP and CH algorithms. We evaluate our FWP-based MMP and CH algorithms on a wide range of large-scale real-world models. Computational results show that our method can improve the speed by a factor of 3-10.

Abstract This paper studies the Voronoi diagrams on 2‐manifold meshes based on geodesic metric (a.k.a. geodesic Voronoi diagrams or GVDs), which have polyline generators. We show that our general setting leads to situations more complicated than conventional 2D Euclidean Voronoi diagrams as well as point‐source based GVDs, since a typical bisector contains line segments, hyperbolic segments and parabolic segments. To tackle this challenge, we introduce a new concept, called local Voronoi diagram (LVD), which is a combination of additively weighted Voronoi diagram and line‐segment Voronoi diagram on a mesh triangle. We show that when restricting on a single mesh triangle, the GVD is a subset of the LVD and only two types of mesh triangles can contain GVD edges. Based on these results, we propose an efficient algorithm for constructing the GVD with polyline generators. Our algorithm runs in O(nNlogN) time and takes O(nN) space on an n‐face mesh with m generators, where N = max{m, n}. Computational results on real‐world models demonstrate the efficiency of our algorithm.