Data intelligence is a pivotal driver of the digital economy, fostering social innovation and high-quality economic growth. Mastering data intelligence equips enterprises with a competitive edge, enabling them to harness technologies, products, and theories to unlock the potential of internal data. The Enterprise Data Intelligence Platform, as a foundational infrastructure, integrates data storage, computation, analysis, and governance, offering end-to-end data life-cycle services. As digital transformation accelerates, building specialized and adaptive data intelligence platforms has become a critical industry focus. This paper explores the evolution, architecture, and implementation of enterprise data intelligence platform, providing insights for enterprise upgrades and envisioning future advancements in the field.

In recent years, with the breakthroughs in generative artificial intelligence technology, AI has become a key approach and effective means for unlocking the value of data elements. The relationship between the big data industry and the AI industry has evolved from a one-way empowerment to a deep integration, giving birth to the data intelligence industry. The collaboration between data and intelligence has together become the core force in transforming the physical world and reshaping the digital world order. The booming development of the data intelligence industry has become the core engine driving innovation and growth in the digital economy. This paper focuses on the topic of data intelligence. It systematically clarifies the technical system of data intelligence, conducts an in-depth analysis of the current status and issues in specific areas of data intelligence, such as data, algorithms, and applications, and also provides an outlook on the future development trends in various fields of the data intelligence industry.

Underwater image enhancement has advanced through deep learning but requires numerous high-quality reference images, which are hard to get. To mitigate this challenge, semi-supervised learning has emerged as a promising approach to minimize the dependency on labeled data. Among the semi-supervised learning methods, the mean teacher method has attracted attention for its potential in enhancing underwater images. However, it is often affected by domain gaps, resulting in the generation of a large number of low-quality pseudo labels, which bias the training results towards the source domain.To address this issue, we introduce a framework named Enhancing Pseudo Label Quality for Semi-Supervised Underwater Image Enhancement (EPLQ-UIE). Specifically, we incorporate adversarial learning and data perturbation within the student model. Domain adversarial learning aligns the distribution of features between the source and target domains, preventing student models from leaning towards source domain data and reducing the impact of domain gaps. Data perturbation enhances the ability of student models to process target domain data by adding target domain features to source domain data, thereby enabling teacher models to generate more reliable pseudo labels. Finally, through comparative experiments on four different underwater image datasets, our method achieved excellent results in both subjective and objective evaluation metrics.

This article presents a multistrategy hybrid dung beetle algorithm (MHDBO) algorithm to tackle the complexities of Autonomous tractor path planning, a problem characterized by intricate optimization objectives and constraints. The proposed MHDBO effectively reduces operational costs, minimizes path lengths, and enhances coverage efficiency in agricultural operations. To improve optimization performance, a novel quantum-assisted offset estimation strategy is introduced, which significantly expands the solution space, improves population diversity, and effectively guides the evolutionary direction of the algorithm to avoid local optima. Additionally, an enhanced double-helix search strategy is designed to strengthen global search capabilities, improving both accuracy and convergence. A reverse learning strategy is integrated, in which a quasi-oppositional learning mechanism is used to avoid large positional deviations typical of traditional reverse learning, thereby maintaining solution stability while enhancing exploration. This helps prevent premature convergence and improves robustness in dynamic environments. To further optimize multitractor coordination, a multistack multitractor scheduling and field path planning model (MSPPM) is developed, addressing the challenges of large-scale agricultural operations. Experimental evaluations show that the proposed MHDBO algorithm achieves a total dispatch path length of 29 093.4007 km, an overall operational cost of 2.622543 million RMB, an execution time of only 4.128759 s, and a memory consumption of$3.588~{\times }~10{^{{9}}}$bytes in the MATLAB environment. Furthermore, MHDBO demonstrates strong performance in coverage, achieving over 95% effective farmland utilization. Compared with state-of-the-art algorithms, including adaptive elite differential evolution algorithm (AEDE), improved 2-opt ant colony optimization (IACO), and adaptive elite chaotic genetic algorithm (AECGA), MHDBO achieves a 10.28% reduction in operational costs, a 5.859% decrease in path length, and executes 15.46% faster, while consuming 16.06% less memory. These results confirm the algorithm’s superiority, efficiency, and scalability in real-time agricultural automation applications.

The concept of "Lakehouse" was proposed by Databricks in 2020. Since "Lakehouse" was first written into Gartner’s Hype Cycle for Data Management in 2021, as a new technology, "Lakehouse" has received unprecedented attention from the enterprises who need digital transformation. More enterprises believe lakehouse is an important infrastructure for digital transformation. Currently, lakehouse is still in its early stage of development. It is not merely a technical research endeavor, but rather a gradual integration of technologies, representing a transitional phase in the evolution of heterogeneous data platform towards integration. This paper focuses on the lakehouse technology, sorts out the development history of the data platform and the practice path of lakehouse technology. It also lists main manufactures and products of lakehouse and provides the judgments for the future development of lakehouse.

The construction of information system stability assurance capabilities has a long history of development. However, with the gradually increasing load pressure of the information system, the distributed architecture has gradually become the mainstream architecture of information system. Furthermore, the stability guarantee of the information system has entered a new stage, which requires extensive optimization in both theoretical and practical aspects. This paper researches on the relevant background, basic principles, key elements, core competencies, and evaluation system of information system stability assurance capacity construction in the new stage. In addition, this paper explores the future development trend of stability assurance work.

With the advancement of digital transformation in various fields, the information industry has developed rapidly in the 21st century. Significant changes in system scale, research and development operation and maintenance models, technical architecture, and user groups have made the impact of information system failures wider and more severe. Disaster recovery systems can protect the data and applications of information systems before disasters occur, ensure the security of data in the event of a disaster, and achieve rapid business recovery. They have become an indispensable foundation for modern data infrastructure. This article provides a brief introduction to various disaster recovery technologies, followed by a summary of the principles that should be followed during the construction process of the disaster recovery system. Finally, it provides a detailed introduction to how to execute and manage the entire process of the disaster recovery system construction and proposes corresponding suggestions.

With the fast development of information technology, the databases used as the fundamental storage and computing component of information system have to deal with much more complicated scenarios with high frequency and concurrency. Distributed databases are becoming more and more common in data-intensive industries such as banking and telecommunications. The selection of distributed database products requires comprehensively consider the function, performance, security, ease of use and stability of the product. The stability of distributed database is known to be difficult in testing and product selection. The method widely used in database testing is TPC-DS (transaction processing performance council-decision support) benchmark, which is only suitable for functional and performance testing. To deal with this drawback, a distributed database stability testing platform based on chaos engineering methodology is developed. Through the perturbation injection by using the stability testing platform, the performance fluctuation of the database under pressure condition can be observed, and the stability of the database can then be evaluated.

This paper reviewed current transactional database benchmark, especially TPC-C. There are some limitations to apply TPC-C into financial industry for benchmarking transactional database. In order to simulate real financial business, the project member has designed a benchmark named Databench-T, which is based on money transfer scenario. The model and workloads of the benchmark is introduced in detail in this paper. The benchmark is implemented by project members to provide a unified toolkit. To run a test, there are four steps: toolkit installation, data generation, workload execute and results display. All those steps are presented in the paper.

With the continuous emergence of cloud computing technology, cloud infrastructure software will become the mainstream application model in the future. Among the databases, relational databases occupy the largest market share. Therefore, the relational cloud database will be the main product of the combination of database technology and cloud computing technology, and will become an important branch of the database industry. This article explores the establishment of an evaluation system framework for relational databases, helping enterprises to select relational cloud database products according to a clear goal and path. This article can help enterprises complete the landing of relational cloud database projects.

In industry, analytic database is known to be difficult in testing and product selection. The traditional TPC-DS (transaction processing performance council-decisioin support) benchmark [1], which only has one retail scene, is not sufficient enough for database product selection of other business type. To deal with this drawback, an analytical database testing platform which can plug-in a variety of business scenes was developed. The telecommunication scene was also provided to explain the design concept, system architecture and scene development of this testing platform. The platform provides scene plug-in function which is easy to use without any programming background. Users can select and edit scenes according to their own business type to meet the requirements of analytical database testing and selection in different industries.