Quick Entry

  • Author Login Author Login
  • Expert Login Expert Login
  • Editorial Office Editorial Office

Quick Icon

WeChat official account
Today's Headlines
  • Latest Articles
  • Online first
  • All Issues
更多
Volume 12 期 5，2026 2026年第12卷第5期

  • TOPIC

    导读 AI导读

    Vol. 12, Issue 5, Pages: 1-2(2026)
      
    14
    |
    23
    |
    0
    <HTML>
    <L-PDF><WORD><XML> ...
    <引用本文> <批量引用> 171921818 false
    更新时间：2026-09-15

  • TOPIC

    Zhang Qunhong, Li Lin, Yang Kun, Chen Shuang, Zhang Binbin, Ye Zishen

    Vol. 12, Issue 5, Pages: 3-20(2026) DOI: 10.11959/j.issn.2096-0271.BDR26084
    摘要：High-quality dataset is a critical foundation for advancing artificial intelligence(AI). However, significant practical challenges were identified in existing development models, including static and preset quality evaluations, difficulties in forming closed-loop business models, and a disconnect from the real economy. In response, a new paradigm for building high-quality dataset was proposed and justified: the development of high-quality dataset was deeply integrated into the data factor market, and data quality was dynamically defined through market-based circulation and multi-dimensional user feedback. This paradigm was centered on“empowerment of the real economy as the main focus and AI training services as a supplement”. Trustworthiness, controllability, and measurability in data circulation were ensured under this paradigm using a “trusted data space based on real-virtual integration”, supported by full-flow circulation mapping and large models for data circulation. On this basis, a full-process technical solution was designed—from data resource market entry to precise distribution—and sustainable business models and incentive mechanisms were established. These efforts provided innovative ideas for shifting the development of high-quality dataset from a “static, supply-driven” approach to a “dynamic, market-driven, and ecosystem-based operational” model.  
    关键词：high-quality dataset;Trusted Data Space;data factors market;dynamic quality evaluation;Reality-Data Integration   
    233
    |
    112
    |
    0
    <HTML>
    <L-PDF><WORD><XML> ...
    <引用本文> <批量引用> 157928993 false
    更新时间：2026-09-15

  • TOPIC

    Cui Wenxuan, Ding Xiaoou, Cheng Haifeng, Qian Zekai, Li Yong, Wang Hongzhi, Wang Chen

    Vol. 12, Issue 5, Pages: 21-42(2026) DOI: 10.11959/j.issn.2096-0271.BDR26172
    摘要：In practical machine learning (ML) model training scenarios, data preparation is still constrained by low resource utilization efficiency, uncontrollable costs, and insufficient optimization under budget constraints. To address these problems, a holistic data preparation framework named CBAprep (cost-benefit-aware data preparation) was proposed, oriented toward cost-benefit collaborative optimization. Mathematical programming was adopted to incorporate data preprocessing, cleaning, transformation, and other operations into a unified cost modeling and decision-making framework, and a cost-benefit analysis and planning method for the overall pipeline was designed. Experimental results showed that CBAprep achieved a better trade-off between cost and performance under limited budget conditions, thereby improving the overall performance and cost-effectiveness of the data preparation process. The results indicate that incorporating cost-benefit analysis and optimization into data preparation strategies facilitates efficient resource allocation and provides useful support for cost control and performance optimization in complex data processing scenarios.  
    关键词：data governance;data preparation;resource scheduling;cost–benefit   
    167
    |
    58
    |
    0
    <HTML>
    <L-PDF><WORD><XML> ...
    <引用本文> <批量引用> 162969639 false
    更新时间：2026-09-15

  • TOPIC

    Zhang Yawei, Tian Liang, Ma Hongwei, Wang Hu, Wang Chenyang, Zha Li

    Vol. 12, Issue 5, Pages: 43-63(2026) DOI: 10.11959/j.issn.2096-0271.BDR26153
    摘要：To address insufficient domain knowledge representation, cross-modal semantic mismatches, and the absence of closed-loop quality governance in multimodal dataset construction for the telecommunications industry, an ontology-driven construction and governance framework, OntoDataset, was proposed. A three-layer architecture for knowledge modeling, semantic reasoning, and data generation was established with a domain ontology as the unified semantic foundation. An ontology-guided algorithm, OntAlign, was also designed to align multimodal data by integrating entity, relation, and contextual constraints. Experiments on real business data from a large telecommunications operator showed that, compared with the best-performing baseline, overall data quality was improved from 86.7% to 93.0%, annotation efficiency was increased from 250 to 455 samples per hour, and the manual intervention rate was reduced from 18% to 12%. In downstream-task evaluations, the AUC for customer churn prediction was 0.897, while the macroaveraged F1 score and accuracy for customer intent recognition were 0.856 and 0.873, respectively. The results indicate that the proposed method improves the quality, consistency, and construction efficiency of multimodal datasets in vertical industries and provides technical support for dataset construction and continuous governance for industry-oriented large models.  
    关键词：multimodal dataset;domain ontology;knowledge graph;cross-modal semantic alignment;data quality governance;telecommunications industry   
    56
    |
    22
    |
    0
    <HTML>
    <L-PDF><WORD><XML> ...
    <引用本文> <批量引用> 171921611 false
    更新时间：2026-09-15

  • TOPIC

    Zhao Congpu, Tan Xutong, Zhu Pujue, Liu Ruiqi, Chen Zheng, Shen Qingni

    Vol. 12, Issue 5, Pages: 64-79(2026) DOI: 10.11959/j.issn.2096-0271.BDR26152
    摘要：High-quality, multimodal data resources serve as the cornerstone for advancing the research and development of medical artificial intelligence models and their clinical translation. However, medical data is characterized by multi-source heterogeneity, sparsity and missing values, privacy sensitivity, and inadequate annotation, all of which directly undermine the generalization capability, interpretability, and clinical applicability of the models. Medical data governance aims to enhance data interoperability, usability, quality controllability, and compliant circulation capacity across the entire data lifecycle. Taking the full lifecycle of medical data as the core thread, this paper systematically reviews the evolutionary trajectory of key governance technologies. First, it examines how multimodal collaborative acquisition and generative data synthesis address the challenge of sample scarcity in the data collection stage. Second, focusing on data standardization and quality improvement, it analyzes the major advances in medical data processing, including semantic transformation, intelligent data cleaning, automated quality assessment, and human-machine collaborative annotation. Third, in response to the demand for cross-institutional data sharing and utilization, it discusses the applications of technologies such as privacy computing, federated learning, and blockchain in compliance-focused governance. Finally, against the backdrop of large-scale model applications, this paper summarizes the practical challenges facing current medical data governance in terms of data bias, model hallucination, and cross-modal alignment, and provides an outlook on the future development trends of data governance.  
    关键词：medical data governance;data quality;data standardization;privacy computation   
    185
    |
    56
    |
    0
    <HTML>
    <L-PDF><WORD><XML> ...
    <引用本文> <批量引用> 162968737 false
    更新时间：2026-09-15

  • TOPIC

    Fu Jia, Chen Yingying, Li Dejun, Chen Xu, Liu Rui

    Vol. 12, Issue 5, Pages: 80-96(2026) DOI: 10.11959/j.issn.2096-0271.BDR26159
    摘要：To address the challenges of diverse airborne data types, difficulties in precise probe alignment, and high data leakage risks during data splitting in weather modification aircraft operations, a multi-source microphysical benchmark dataset oriented toward data governance and leakage-controlled evaluation was constructed. Through methods including dirty data rejection, adaptive parsing of multi-character encodings, 1 Hz fusion anchored by the DXAS probe, and non-negative physical ice thickness generation, airborne data from 43 flights were transformed into a high-quality fused table containing 336 570 samples. Furthermore, a strict nested leave-one-session-out (LOSO) cross-validation protocol was established to block event-level leakage. Benchmark testing using the XGBoost algorithm verified the validity of icing onset signals in the airborne data and established a baseline for regression tasks under leakage control. This dataset provides a standardized data foundation for aircraft icing mechanism analysis and early-warning algorithm development, offering a rigorous evaluation reference for enhancing risk monitoring and ensuring flight safety.  
    关键词：weather modification;airborne dataset;dataset construction;data governance;second-level alignment;leakage control   
    106
    |
    59
    |
    0
    <HTML>
    <L-PDF><WORD><XML> ...
    <引用本文> <批量引用> 157454736 false
    更新时间：2026-09-15
查看更多

Hotlights

更多

Data governance framework for artificial intelligence

Data plays a crucial role in the development and application of artificial intelligence, which has become a consensus in industry and academia. Based on the interactive relationship between artificial intelligence and data, as well as the data-centric AI development practice, this paper proposes a data governance framework for artificial intelligence, which consists of six stages: source data governance, pre-trained data governance, evaluation data governance, fine-tuning data governance, inference data governance, and operation and maintenance data governance. Each stage has its key tasks and technologies. At the same time, this paper deeply analyzes the data governance cases and successful experiences of ChatGPT, Ziya2 and artificial intelligence models in the energy field to verify the effectiveness of the framework. The result shows that the framework plays an important role in improving the performance of artificial intelligence models and optimizing data governance processes. The framework provides reference for theoretical and practical innovation of data governance oriented to artificial intelligence.

> COLUMN

LI Jifeng, ZHANG Chenglong, LIU Xin, CHEN Jinyu, ZHANG Jinming, BI Chao

Data governance framework for artificial intelligence
0