While person re-identification (Re-ID) has achieved remarkable progress in controlled laboratory settings, its widespread deployment in real-world urban surveillance remains impeded by the significant domain gaps caused by environmental dynamics and sparse camera topologies. In such operational scenarios, relying solely on visual appearance leads to severe visual ambiguity, while the continuous expansion of camera networks induces catastrophic forgetting. To bridge this “lab-to-real” gap, we present a robust, street-surveillance-oriented Re-ID solution validated across a diverse array of benchmarks. Our framework incorporates three key innovations: (1) We construct a deployable Teacher–Student framework to ensure stable feature transfer from labeled source domains to noisy target environments, enabling robust adaptation without manual supervision; (2) Addressing system sustainability, we design a diversity-preserving Dynamic Data Replay mechanism based on Farthest Point Sampling (FPS) to prevent catastrophic forgetting as the network continuously expands; (3) A Spatiotemporal Feature Fusion (STFF) module is developed to resolve visual ambiguity in sparse networks by imposing explicit physical constraints to filter out spatiotemporally infeasible candidates. Extensive evaluations on a proprietary benchmark derived from the live Hainan surveillance system and a series of mainstream datasets demonstrate that our method significantly outperforms existing approaches, offering a superior solution for practical deployment in smart city infrastructure.

For languages with no annotated resources, transferring knowledge from rich-resource languages is an effective solution for named entity recognition (NER). While all existing methods directly transfer from source-learned model to a target language, in this paper, we propose to fine-tune the learned model with a few similar examples given a test case, which could benefit the prediction by leveraging the structural and semantic information conveyed in such similar examples. To this end, we present a meta-learning algorithm to find a good model parameter initialization that could fast adapt to the given test case and propose to construct multiple pseudo-NER tasks for meta-training by computing sentence similarities. To further improve the model's generalization ability across different languages, we introduce a masking scheme and augment the loss function with an additional maximum term during meta-training. We conduct extensive experiments on cross-lingual named entity recognition with minimal resources over five target languages. The results show that our approach significantly outperforms existing state-of-the-art methods across the board.

To better tackle the named entity recognition (NER) problem on languages with little/no labeled data, cross-lingual NER must effectively leverage knowledge learned from source languages with rich labeled data.Previous works on cross-lingual NER are mostly based on label projection with pairwise texts or direct model transfer.However, such methods either are not applicable if the labeled data in the source languages is unavailable, or do not leverage information contained in unlabeled data in the target language.In this paper, we propose a teacher-student learning method to address such limitations, where NER models in the source languages are used as teachers to train a student model on unlabeled data in the target language.The proposed method works for both single-source and multi-source crosslingual NER.For the latter, we further propose a similarity measuring method to better weight the supervision from different teacher models.Extensive experiments for 3 target languages on benchmark datasets well demonstrate that our method outperforms existing state-of-theart methods for both single-source and multisource cross-lingual NER.

Prior work in cross-lingual named entity recognition (NER) with no/little labeled data falls into two primary categories: model transfer- and data transfer-based methods. In this paper, we find that both method types can complement each other, in the sense that, the former can exploit context information via language-independent features but sees no task-specific information in the target language; while the latter generally generates pseudo target-language training data via translation but its exploitation of context information is weakened by inaccurate translations. Moreover, prior work rarely leverages unlabeled data in the target language, which can be effortlessly collected and potentially contains valuable information for improved results. To handle both problems, we propose a novel approach termed UniTrans to Unify both model and data Transfer for cross-lingual NER, and furthermore, leverage the available information from unlabeled target-language data via enhanced knowledge distillation. We evaluate our proposed UniTrans over 4 target languages on benchmark datasets. Our experimental results show that it substantially outperforms the existing state-of-the-art methods.

Generally speaking, sentiment lexicons employed in the majority of current sentiment analysis systems are trained globally from public data stream source or other large independent corpus. However, sentiments are rather subjective and personal states of mind that the individuality and diversity of characteristics, particular writing habit and idiolect could play a crucial role in the judgment of sentiment expressed by a specific user. In this paper, we present a novel feature construction method to combine user-based and global lexicon features in sentiment analysis for short social media text. After the creation of user-based sentiment lexicons from user-timeline corpus, a rule-based fusing approach is adopted subsequently to generate user-based lexicon features in combination with general lexicon features. Experiments show that user-based features may capture potential user preferences hence adjusting the bias caused by representing an individual's sentiment with an averaged lexicon score, and our proposed method yield better results in comparison with some of the state-of-the-art sentiment analysis systems in twitter.

In recent years, the service system based on taxi GPS trajectory gradually becomes a hot research topic. In this paper we first give some descriptions and definitions of the taxi GPS trajectory problems. Different from the traditional recommendation system, the service system based on taxi GPS trajectory will lead to some special challenges and we will give corresponding solutions to solve these problems. Therefore, we propose a recommendation system framework for this issue via the emphasis on temporal and spatial information mining. Then, we discuss the different classification method by different points of views including the statistics of spatial information, the modeling of time information, mining methods and knowledge discovery models. Finally, we point out the promising directions in this field.

To confirm the ingenuity of a patent, the drawings appeared in the patent document play a great role in the comparison of similar patent and can further combine with text-based image retrieval for accurate search. Considerable work has been done in image retrieval using shape, color and texture information. However, patent images are usually binary with complex shapes, no color and little texture information, thus little effort has been made specifically for patents. In this paper, we proposed a novel method named the hierarchical oriented gradient histogram, which extracts the local and global gradient distribution of the image. It can be used in binary patent images which are very complex and cannot be easily segmented into shapes. Experiments on a public database demonstrated that the proposed algorithm could get higher accuracy than other state-of-the-art approaches. Because the dimension of feature of an image is less than 200, our method can be utilized for patent image retrieval in real-time.

In this paper, we proposes a method which use locality-constrained linear coding (LLC) and spatial pyramid matching (SPM) for patent image classification. Patent images usually have no texture and color information which makes it hard for recognition. Many methods based on contour, shape or edge of image have been proposed, however, as far as we are concerned, our method is the first attempt using coding features for patent image classification. First, we extract dense Scale-Invariant Feature Transformation (SIFT) features and use k-means clustering to train a codebook which based on LLC. Second, we divide the image into increasing fine sub-regions and generate the feature for each sub-region as SPM do. Finally, we use a linear SVM classifier for patent image classification. The experiment on a public database for patent image has demonstrated our method has achieved the stated-of-the-art accuracy rate of 94.2%. It proves model based on SPM and LLC have bright future in patent image recognition.

Nowadays, wind power has become the focus of development. Under the background of large scale wind farm construction, this paper starts from the importance of wind farm project design and seeks to describe and analyze the current situation and problems of the wind farm design management. In addition, the paper proposes a full life-cycle model of wind farm engineering design, based on which, presents a knowledge management service model as well. On this basis, it designs and develops a management information system for the owner and the other participating units, providing design management, quality management, document management service, and so on. The system provides information management and knowledge management service, so as to enhance the quality of wind farm construction and contribute to the development of wind power industry.

As a new management pattern, `cluster supply chain' (CSC) can help Small and Medium Enterprises (SMEs) to face the global challenges through all kinds of collaboration. However, a major challenge in implementing CSC is the gap between the related theories and practices in the field. The recent rapid commercialization and adoption of `service' technologies have driven a process of transforming theories into practices. In an effort to provide a better understanding of this emerging phenomenon, this paper presents the implementation process of CSC in the context of JingCheng Mechanical \& Electrical Holding co., ltd.(JCH) as a case study. The cast study of JCH suggests that the key problems in the practice of cluster supply chain: When do small firms use cluster supply chain to do business? How do small firms use cluster supply chain? Only after clarifying those problems, the actual construction and operation of cluster supply chain does show successful results as it should be.

As a new type of management pattern, ”cluster supply chain” (CSC) can help SMEs to face the global challenges through all kinds of collaboration. However, a major challenge in implementing CSC is the gap between theory and practice in the field. The recent rapid commercialization and adoption of ”service” technologies has driven a process of transforming theory into practices. In an effort to provide a better understanding of this emerging phenomenon, this paper presents two key elements of cluster supply chain, including: an instanced model for cluster supply chain alliance (CSA), and the architecture of service supporting system. Cluster supply chain has propelled a structural change in buyers-intermediaries-sellers relationships, and accelerated internationalization of small- and medium-sized enterprises.

The logistics information system usually includes many subsystems with close interaction. The cooperative design is an inevitable in the development issue. This paper adopts the simple factory pattern to solve the programming problem at the beginning of system design that every subsystem is defined with interfaces to its referencing subsystems in order to invoke during programming and reduce the coupling with other subsystems. When adopting this simple factory pattern, it is convenient to design and develop the overall system cooperatively, and saving time and improving efficiency simultaneously.

With the Increased competition in the transport market, the new international logistics service provider must develop solutions tailored to meet the unique and special needs of each customer. Building upon the foundation of 3rd party outsourcing, 4PL emerges and acts as an electronic intermediator that connects suppliers, producers, carriers and customers to dynamic supply chain networks. Of such transportation networks the operative planning task of the hub-and-spoke network is a challenging task for the management. In particular, the transport management has to decide where all quantifies within the transportation network flow over the hub from or to the depots, or whether a hybrid hub-and-spoke network is preferred in which direct transports also take place. Upon these, Optimization model is studied and a 4PL optimization system for these operative planning tasks is developed and applied to a real case of a Chinese logistic service provider. The result of the case indicates that the model and algorithm presented in the paper can provide a reasonable supply chain scheduling scheme.

The paper presents a study on the operation management of virtual enterprises. A virtual enterprise is a temporary alliance of member enterprises formed to exploit fast-changing opportunities. During the operation of a virtual enterprise, its constituent member enterprises, which are geographically distributed and organizationally independent, collaborate with each other to execute the whole business process of the virtual enterprise. Therefore, how to support the business process oriented virtual enterprise cooperative operation is the focus of our study on the operation management of virtual enterprises. In order to achieve this, we propose a scheme that consists of three parts: the modeling of the distributed business process in virtual enterprises, the modeling of virtual enterprises themselves, and the control of the business process execution in virtual enterprises based on the two kinds of models established in the former two parts.

Based on the object-oriented approach and Interval Timed Coloured Petri nets (ITCPNs), we study the modeling of distributed business processes in virtual enterprises. We model each member enterprise in a virtual enterprise as an object, and define both the internal model of the member enterprise object and the interaction model among the member enterprise objects based on ITCPNs. These two kinds of models are two views of the system and when combined, a complete model for the whole business process of the virtual enterprise is achieved.

We propose an agent-based virtual enterprise model and provide the agent collaboration mechanisms under the model, thereby achieving agent-based virtual enterprise modeling and operation control. We establish a network of distributed activity agents and resource agents for each member enterprise, collaborations among which can deal with activity precedence arrangements and activity-resource assignments in each sub-business-process. Then, we model each member enterprise in the virtual enterprise as an agent. Collaborations among member enterprise agents can deal with activity coordination and resource sharing among various sub-business-processes. Thus, our agent-based approach achieves distributed control over the whole business process execution of the virtual enterprise.

The current network environment cannot support the creation and operation of virtual enterprises effectively. The paper analyzes characteristics of virtual enterprises, including information, communication, interaction, and the existing network environment, and then presents an electronic market architecture for virtual enterprises, which is a peer-to-peer network (P2P-VEN, Peer-to-Peer Virtual Enterprise Network). We discuss the core technologies of P2P-VEN, including communication, transfer, search and security. P2P-VEN supports the creation and operation of virtual enterprises, and provides effective information sharing and exchange. It ensures autonomy and privacy of individual enterprises, and is appropriate as the network infrastructure of virtual enterprises.

Partner selection is an important and complex decision problem for virtual enterprises (VEs). In this paper, we present a goal ontology for partner selection in a VE, and provide self-contained computing theory and a problem solving meta-model to manage and analyze VE goals. The paper focuses on the concept, attributes and constraints of goal ontology, decomposition relationship and redundancy of the goal system, and changeability and consistency of constraints.

Partner selection is an important and complex decision problem for virtual enterprises. We present the Tenderee Decision Support System based on goal ontology, which adopts distributed constraint satisfaction and multi-attribute utility theory to solve and optimize the problem, and provides powerful decision support capability by human-machine interaction in the aspects of setup and maintenance of goal, selection and optimization partner.

An enterprise should know its own competence sufficiently, use and improve it reasonably and continually. In this paper, we present competence ontology of manufacturing enterprise, which provides meta-model and self-contained computing micro-theory for modeling and analysis of competence of manufacturing enterprise. The paper focuses on the competence problems, concept, properties and constraints of competence ontology.

The paper discusses a decision support system (DSS) for bidding by manufacturing enterprises. It is based on a competence ontology, which can evaluate the competence of an enterprise, analyse its feasibility and prospective benefits, and optimize bids. The ontology of enterprise competence provides a terminology and micro-theory for the description of and reasoning about enterprise competence. The DSS adopts a scheduling system based on multi-agent negotiation, which estimates the time and cost to achieve the goal of inviting public bidding, and which provides an important basis for enterprise bidding decisions.

The authors propose an agent based approach for workflow management, aiming at achieving flexible and dynamic workflow management in the distributed environment. We treat agents as autonomous entities with abilities to solve problems independently and propose an agent hierarchy as well as an agent model. The agent hierarchy consists of three kinds of agents at three neighboring levels: activity agents, role agents and actor agents. With our agent hierarchy, the gap between the dynamic workflow enactment and the static business process model can be bridged. The three levels of agents collaborate with each other through the contract-net approach and the feedback mechanism. Through such collaboration, activity-actor assignments can be achieved flexibly and dynamically; and the workflow execution can be monitored and managed in a real time fashion. The agent model consists of three sub-agents: a message-receiving one, a decision-making one and a message-sending one. Connected by two message queues, these three collaboratively operating sub-agents can complete functions of an agent effectively. Through message transfers of the message-receiving/sending sub-agents and the message based behavior-performing of the decision-making subagent, individual agents as well as the agent hierarchy can operate effectively and efficiently.

The paper presents our studies on Petri net based business process modeling for virtual enterprises. A virtual enterprise (VE) is a temporary alliance of member enterprises. Generally, a business process consists of a series of logically interrelated activities, to which appropriate resources will be assigned. In the virtual enterprise, both the activities and the resources are in multiple member enterprises, while collaborations exist among these distributed activities and resource sharing is often required. Thus, the modeling of business processes in virtual enterprises is challenging work. Based on Petri nets, the paper proposes a business process modeling method for virtual enterprises, which focuses on the two basic elements of the business process: activities and resources, while addressing their properties of distribution and collaboration particular to virtual enterprises.

When faced with a fast-changing global market, an enterprise needs to understand its own competence clearly, as well as assess its competence correctly according to the market demands. This paper focuses on the matchmaking of enterprise competence with competence requests. We propose a kind of knowledge-based decision support system (DSS) for the problem of competence matchmaking, which uses a multi-agent system. Agents in the system can be requestors or providers of competence, which can take advantage of the knowledge in the competence knowledge base to settle the problem of competence matchmaking through reasoning and interaction with other agents. This DSS has the capability of automatic quantitative reasoning, as well as its knowledge-based decision support capability.

It is very important for an enterprise to understand its competence and to improve it continuously, so as to better engage with the competitive market environment. This enterprise competence information is also valuable for potential cooperation among different enterprises. However, this information is often not available even within the enterprise itself, let alone for other enterprises, because there is no set of general methods for the modeling and management of enterprise competence. In this paper, we present a competence management system based on a common enterprise competence model (ECM), including a competence modeling method, system architecture, model-based decision support, and a core technology for system implementation.

When modeling a scheduling problem by Petri net most researchers used the so-called pure net. The net can not deal with the allocations of time resources and the critical resources such as processors. These resources should be the elements of the net, so in the strict sense, the net is not a timed Petri net. In this paper, we present an extended Petri net model, a scheduling model with time resources ( called TREM net). TREM net can model both processors and times so it can easily treat the allocations of both simultaneously. We discuss in details the method of solving T-invariants of TREM net and the close relation between T-invariants and scheduling. The idea of combining AI technique with Petri net based scheduling model is also presented in the paper.>