Zhejiang University

Zhejiang Lab

Shuiguang Deng

Jianwei Yin

Ying Li

Zhaohui Wu

Jian Wu

Cheng Zhang

Yingchun Yang

Shengye Pang

Bangpeng Zheng

Zhiling Luo

Gang Pan

Huan Zhang

Ge Su

Yongheng Shang

Maolin Zhang

Wenqiu Zeng

Yucheng Guo

Liye Cheng

Xuhong Zhang

Chen Zhi

Meilin Liu

Xinkui Zhao

Yuxiang Cai

Meng Xi

Yihua Mao

Min Kang

Pan Lv

Fengjuan Wang

Yishan Chen

Jiayin Luo

Jintao Chen

Disclosed in the present disclosure is a low-cost and zero-shot online log parsing method based on a large language model, including: firstly, extracting content of a log in a log message using regular expressions, then, performing regular expression matching with a log template in a database; if the matching is successful, updating a log sample corresponding to the log template; if the matching fails, conducting a dialogue with the large language model to obtain a new log template; performing template correction to prevent the log template generated by the large language model from being incapable of performing regular expression matching with the log message; performing template merging when a new template is generated; performing template splitting when the log sample is updated; and for all log templates to be added to the database, firstly, normalizing the log templates by post-processing, and then storing the log templates to the database. The log template generated by the present disclosure outperforms conventional methods in terms of word parsing accuracy, with significant advantages in speed and cost compared with the direct use of the large language model to perform log parsing tasks.

Low-cost and zero-shot online log parsing method based on large language model

Disclosed in the present invention is a method for dynamic evolution of a composite service, which belongs to the field of service composition in software engineering, and comprises a composite service dynamic evolution process, a distributed service evolution mechanism and a multi-layer process evolution mechanism. The present invention can dynamically adjust the composite service according to a real-time state of an atomic service under a condition of dynamic change of the service quality of the atomic service, select different candidate services, and avoid falling into a local optimal solution. At the same time, the present invention can dynamically optimize a service process according to the selection of the atomic service under the condition of loose coupling and low constraints, while taking into account the evolution of a single service and the co-evolution of services. Further, in order to reduce the load of the atomic service, the present invention utilizes service performance data generated when the service is actually running, so as to avoid empty requests due to obtaining the service quality information of the atomic service.

Method for dynamic evolution of composite service

The present invention discloses a semantic segmentation method for cross-satellite remote sensing images based on unsupervised bidirectional domain adaptation and fusion. The method includes training of bidirectional source-target domain image translation models, selection of bidirectional generators in the image translation models, bidirectional translation of source-target domain images, training of source and target domain semantic segmentation models, and generation and fusion of source and target domain segmentation probabilities. According to the present invention, by utilizing source-target and target-source bidirectional domain adaptation, the source and target domain segmentation probabilities are fused, which improves the accuracy and robustness of a semantic segmentation model for the cross-satellite remote sensing images; and further, through the bidirectional semantic consistency loss and the selection of the parameters of the generators, the influence due to the instability problem of the generators in the bidirectional image translation models is avoided.

Semantic segmentation method for cross-satellite remote sensing images based on unsupervised bidirectional domain adaptation and fusion

A method and a device of intra-chip routing of neural tasks for an operating system of a brain-inspired computer are provided. The method includes determining an area defined by target cores, and determining target cores in a row furthest from an edge routing area; determining whether the target cores need to be configured with relay routing; searching nearest edge routing cores in the edge routing area for all the target cores in the area defined by the target cores; configuring the target cores in a far-to-near principle; and searching relay routing cores and the nearest edge routing cores by a shortest path manner and a maximum step length of a single routing manner.

Device and method of intra-chip routing of neural tasks for operating system of brain-inspired computer

Disclosed in the present invention is a spontaneous edge application deployment and pricing method based on an incentive mechanism. The method comprises the following steps: building an edge end application oriented spontaneous deployment system architecture; then proposing an incentive mechanism aiming at spontaneous edge application deployment and prizing; solving the spontaneous edge application deployment and prizing problem based on a backward induction method, thereby obtaining an optimal deployment solution of an edge server and an optimal prizing strategy of an application provider.

Spontaneous edge application deployment and pricing method based on incentive mechanism

The present invention discloses a fair task offloading and migration method for edge service networks, taking the Pareto optimality of the utility function of all user tasks executed by the edge system as the optimization objective, this approach not only takes into account the constraints of edge network resources, but also ensures the maximization of the utility function of all user tasks in the system, it proposes a new quantitative measurement index for improving the task utility quality under multi-user competition. In addition, the present invention uses the graph neural network and reinforcement learning algorithm to solve the final optimization goal, this algorithm has high execution efficiency and returns accurate approximate results, which is particularly suitable for the scene of edge network system under multi-user complex tasks, so that when multi-user tasks compete for network resources, the edge computing network system can efficiently obtain the Pareto optimal result of multi-user utility function, greatly improving the service quality and user experience of edge network environments.

Fair task offloading and migration method for edge service networks

Disclosed is a service quality prediction method in a service network environment. The method includes the steps of (1) constructing a service uncertainty quality model of a user through service quality log information generated when the user invokes the same or different services multiple times; (2) tapping a set of similar users of a target user based on the service uncertainty quality model of the user; and (3) improving a matrix factorization algorithm based on similarity information of the set of similar users of the target user to achieve accurate prediction of service quality. The service quality prediction method of the present invention can provide a basis for decision-making, perform targeted service optimization based on the service quality information, and provide more accurate service recommendation.

Method for predicting Quality of Service in service network environment

The present invention discloses a network burst load evacuation method for edge servers, which takes a time and average penalty function of all tasks performed by the edge system as a minimum optimization goal. This method not only takes into account the fairness of all users in the system, but also ensures that the unloading tasks of all users in the system can be completed in a relatively shortest time, and a new quantitative measure is proposed for improving user QoS response. In the implementation process of the algorithm in the present invention, a particle swarm algorithm is used to solve an optimal target of the system, This algorithm has a fast execution speed and high efficiency, and is especially suitable for a scene of an edge computing network system, so that when a sudden load occurs, an edge computing network system can respond in a very short time and complete the evacuation of the load, which greatly improves the fault tolerance and stability of the edge network environment.

Network burst load evacuation method for edge servers

The present invention discloses a service caching method for a cross-border service network, wherein the method includes: a cache space of a service switch node is divided into a resident area, a change area, a pre-reclaimed area and a maintenance index area; among them, a cache hit frequency is: a resident area&#x3e;a change area&#x3e;a pre-reclaimed area, and the maintenance index area is used for separate storage services call path. when a service call is generated, a cache content in the cache space is replaced according to a cache value of a missed cache or a hit cache; a service router and service switch nodes in the corresponding area jointly form a hierarchical cache mode. When the cache space of any node in the service switch node is insufficient, the service switch nodes in the same area perform collaborative cache and store them in other cache space of the service switch node through indexing. The method provided by the present invention can improve the cache utilization efficiency in the cross-border service network, thereby accelerating service invocation.

Service caching method for a cross-border service network

The present disclosure discloses an unsupervised domain adaptation method, a device, a system and a storage medium of semantic segmentation based on uniform clustering; first, a prototype-based source domain uniform clustering loss and an empirical prototype-based target domain uniform clustering loss are established, to reduce intra-class differences of pixels responding to the same category; meanwhile, the pixels with similar structures but different classes are driven away from each other, wherein they tend to be evenly distributed, increasing the inter-class distance and overcoming the problem that the category boundaries are unclear during the domain adaptation process; next, the prototype-based source domain uniform clustering loss and the empirical prototype-based target domain uniform clustering loss are integrated into an adversarial training framework, which reduces the domain difference between the source domain and the target domain, thus improving the accuracy of semantic segmentation.

Unsupervised domain adaptation method, device, system and storage medium of semantic segmentation based on uniform clustering

A method and system of graph feature extraction and graph classification based on adjacency matrix is provided. The invention first concentrates the connection information elements in the adjacency matrix into a specific diagonal region of the adjacency matrix which reduces the non-connection information elements in advance. Then the subgraph structure of the graph is further extracted along the diagonal direction using the filter matrix. Further, it uses a stacked convolutional neural network to extract a larger subgraph structure. On one hand, it greatly reduces the amount of computation and complexity, getting rid of the limitations caused by computational complexity and window size. On the other hand, it can capture large subgraph structure through a small window, as well as deep features from the implicit correlation structures at both vertex and edge level, which improves speed and accuracy of graph classification.

System and method of graph feature extraction based on adjacency matrix

Disclosed is system and method of connection information regularization, graph feature extraction and graph classification based on adjacency matrix. By concentrating the connection information elements in the adjacency matrix into a specific diagonal region of the adjacency matrix in order to reduce the non-connection information elements in advance. The subgraph structure of the graph is further extracted along the diagonal direction using the filter matrix. Then a stacked convolutional neural network is used to extract a larger subgraph structure. On the one hand, it greatly reduces the amount of computation and complexity, solving the limitations of the computational complexity and the limitations of window size. And on the other hand, it can capture large subgraph structure through a small window, as well as deep features from the implicit correlation structures at both vertex and edge level, which improves the accuracy and speed of the graph classification.

System and method of connection information regularization, graph feature extraction and graph classification based on adjacency matrix

A method for semantic service registration and query based on WordNet is disclosed. The method includes the following steps: (1) semantic service registration: a service provider registers a service and uploads the Web Service Description Language (WSDL) document corresponding to the service, and a system parses the WSDL document to form a service description tree, then constructs a WordNet ontology tree according to the input of the service, performs a semantic annotation on the input/output of the service to form a Web Service Semantic Description Document (WSDL-S), and finally stores it in a register library; (2) semantic service discovery: a service requester inputs the information of service type, semantic information of the service input/output and other user-defined information to the register library to retrieve the services meeting the requirements; and (3) similarity sorting: the services meeting a certain threshold are sorted in descending order. The method has the advantages of the combination of WordNet ontology library and the semantic description language of WSDL-S, and definite semantic meaning.

Growing community of inventors

Hangzhou, China

Shuiguang Deng