Ranking relevance is a fundamental task in e-commerce search, directly affecting ranking quality and consumer experience. Although inherently an ordinal classification problem, it is commonly formulated as conventional multi-class classification, which overlooks the natural order among relevance levels and assigns equal penalties to adjacent and distant misclassifications. This mismatch leads to suboptimal learning objectives for practical relevance evaluation. To address this issue, we propose a unified cascaded binary classification framework applicable to both large language model inference and online BERT-based inference, which reformulates relevance estimation as a sequential decision process and decomposes multi-class prediction into a series of ordered binary judgments from higher to lower relevance tiers. For large language models, we design a step-wise reasoning procedure with pruning strategies and tier-specific reward functions. For the online BERT model, we replace the conventional classification head with multiple level-wise binary classifiers and distill the capabilities of large language models into the online model. Extensive offline industrial benchmark evaluations and online A/B experiments demonstrate that the proposed framework substantially improves relevance performance, reducing the online bad-case rate by 15.94\%. Further analyses suggest that tier-wise modeling is effective for relevance estimation.

Training data pre-processing pipelines are essential to deep learning (DL). As the performance of model training keeps increasing with both hardware advancements (e.g., faster GPUs) and various software optimizations, the data pre-processing on CPUs is becoming more resource-intensive and a severe bottleneck of the pipeline. This problem is even worse in the cloud, where training jobs exhibit diverse CPU-GPU demands that usually result in mismatches with fixed hardware configurations and resource fragmentation, degrading both training performance and cluster utilization. We introduce GoldMiner, an input data processing service for stateless operations used in pre-processing data for DL model training. GoldMiner decouples data pre-processing from model training into a new role called the data worker. Data workers facilitate scaling of data pre-processing to anywhere in a cluster, effectively pooling the resources across the cluster to satisfy the diverse requirements of training jobs. GoldMiner achieves this decoupling in a fully automatic and elastic manner. The key insight is that data pre-processing is inherently stateless, thus can be executed independently and elastically. This insight guides GoldMiner to automatically extract stateless computation out of a monolithic training program, efficiently disaggregate it across data workers, and elastically scale data workers to tune the resource allocations across jobs to optimize cluster efficiency. We have applied GoldMiner to industrial workloads, and our evaluation shows that GoldMiner can transform unmodified training programs to use data workers, accelerating individual training jobs by up to 12.1x. GoldMiner also improves average job completion time and aggregate GPU utilization by up to 2.5x and 2.1x in a 64-GPU cluster, respectively, by scheduling data workers with elasticity.

The Graph Convolutional Network (GCN) and its variants are powerful models for graph representation learning and have recently achieved great success on many graph-based applications. However, most of them target on shallow models (e.g. 2 layers) on relatively small graphs. Very recently, although many acceleration methods have been developed for GCNs training, it still remains a severe challenge how to scale GCN-like models to larger graphs and deeper layers due to the over-expansion of neighborhoods across layers. In this paper, to address the above challenge, we propose a novel layer-wise sampling strategy, which samples the nodes layer by layer conditionally based on the factors of the bi-directional diffusion between layers. In this way, we potentially restrict the time complexity linear to the number of layers, and construct a mini-batch of nodes with high local bi-directional influence (correlation). Further, we apply the self-attention mechanism to flexibly learn suitable weights for the sampled nodes, which allows the model to be able to incorporate both the first-order and higher-order proximities during a single layer propagation process without extra recursive propagation or skip connection. Extensive experiments on three large benchmark graphs demonstrate the effectiveness and efficiency of the proposed model.

Graph Convolutional Networks (GCNs) achieved tremendous success by effectively gathering local features for nodes. However, commonly do GCNs focus more on node features but less on graph structures within the neighborhood, especially higher-order structural patterns. However, such local structural patterns are shown to be indicative of node properties in numerous fields. In addition, it is not just single patterns, but the distribution over all these patterns matter, because networks are complex and the neighborhood of each node consists of a mixture of various nodes and structural patterns. Correspondingly, in this paper, we propose Graph Structural-topic Neural Network, abbreviated GraphSTONE(1), a GCN model that utilizes topic models of graphs, such that the structural topics capture indicative graph structures broadly from a probabilistic aspect rather than merely a few structures. Specifically, we build topic models upon graphs using anonymous walks and Graph Anchor LDA, an LDA variant that selects significant structural patterns first, so as to alleviate the complexity and generate structural topics efficiently. In addition, we design multi-view GCNs to unify node features and structural topic features and utilize structural topics to guide the aggregation. We evaluate our model through both quantitative and qualitative experiments, where our model exhibits promising performance, high efficiency, and clear interpretability.

To depict ubiquitous relational data in real world, network data have been widely applied in modeling complex relationships. Projecting vertices to low dimensional spaces, quoted as Network Embedding, would thus be applicable to diverse real-world predicative tasks. Numerous works exploiting pairwise proximities, one characteristic owned by real networks, the clustering property, namely vertices are inclined to form communities of various ranges and hence form a hierarchy consisting of communities, has barely received attention from researchers. In this paper, we propose our network embedding framework, abbreviated SpaceNE, preserving hierarchies formed by communities through subspaces, manifolds with flexible dimensionalities and are inherently hierarchical. Moreover, we propose that subspaces are able to address further problems in representing hierarchical communities, including sparsity and space warps. Last but not least, we proposed constraints on dimensions of subspaces to denoise, which are further approximated by differentiable functions such that joint optimization is enabled, along with a layer-wise scheme to alleviate the overhead cause by the vast number of parameters. We conduct various experiments with results demonstrating our model's effectiveness in addressing community hierarchies.

The master-slave control mode of minimally invasive surgery (MIS) robot systems improves intuitive manipulability of MIS, but it cuts off direct force feedback between the master and slave manipulator. In order to design a MIS robot with both intuitive manipulability and direct force feedback, the following studies were conducted. Firstly, based on a new task division of the console surgeon and operating-table surgeon, a new model of an MIS robots system was proposed, in which the operating-table surgeon worked as main surgeon and the console surgeon worked as assistant surgeon. Secondly, according to the workspace requirement of MIS, a symmetrical cable-driven MIS robot was designed including structure, layout of driving cable and motion characteristics. Thirdly, the prototype was manufactured and related experiments were carried out, such as flexibility experiment, position control accuracy experiment and motion performance experiment. The study is important for both improving intuitive manipulability and realizing force feedback of MIS robot.

Network embedding is a method to learn low-dimensional representation vectors for nodes in complex networks. In real networks, nodes may have multiple tags but existing methods ignore the abundant semantic and hierarchical information of tags. This information is useful to many network applications and usually very stable. In this paper, we propose a tag representation learning model, Tag2Vec, which mixes nodes and tags into a hybrid network. Firstly, for tag networks, we define semantic distance as the proximity between tags and design a novel strategy, parameterized random walk, to generate context with semantic and hierarchical information of tags adaptively. Then, we propose hyperbolic Skip-gram model to express the complex hierarchical structure better with lower output dimensions. We evaluate our model on the NBER U.S. patent dataset and WordNet dataset. The results show that our model can learn tag representations with rich semantic information and it outperforms other baselines.

The ＂2011·3·25＂ Event in Suijiang,Yunnan is a typical group event in reservoir resettlement,with the nature being interest conflict.The main causes are inconsiderate thoughts beforehand and unclear compensation afterwards.The right to participate and involuntary resettlement should be given to people in order to settle this kind of group events.

Deterministic replay is a powerful approach for debugging multithreaded and distributed applications that involve complex nondeterminism. Recently, several replay tools based on user-space libraries were proposed. Such library-based tools are lightweight and efcient, and focus on the application being debugged. However, several signicant problems remain, such as in-address-space tool interference, absence of support for advanced features in modern operating systems (e.g. asynchronous I/O), and error-prone engineering efforts for wrapping a large set of API functions to enable logging and replaying. We propose three mechanisms to cope with these problems. First, we resolve the tool interference problem by space separation, using a virtual execution layer composed of intercepted functions. The virtual execution layer isolates both memory and execution of the application from supporting libraries and in-address-space replay tool, hence systematically removes the interference. Second, based on an event-view of the execution, we are able to handle advanced operating features such as asynchronous I/O and exceptional control o w (e.g. signal). Third, instead of manually coding the wrappers, we use an annotation-aware code generator so that the process of producing the wrappers is more robust and efcient. We have handled more than 750 API functions and the code generator is responsible for more than 37,000 lines of wrapper code. To the best of our knowledge, these three mechanisms, space separation, event-based replay, and annotation-aware code generation are novel for library-based deterministic replay. We have successfully replayed many large and complex software packages with low overhead, including both server and distributed system, including those developed by ourselves and by the community. MySQL and Apache, for instance, experienced around 10% slowdown.