半监督学习下医疗术语信息关键词模糊检索算法

Semi-Supervised Learning-Based Fuzzy Retrieval Algorithm for Medical Term Information Keywords

摘要

医疗领域术语具有强异构性, 且标注数据稀缺, 导致语义特征表达较弱, 进而使检索精度较低, 为此, 提出了半监督学习下医疗术语信息关键词模糊检索算法。利用词向量对原始医疗术语信息进行标准化处理与映射, 基于核心元数据信息节点和用户查询信息节点, 确定检索样本之间的置信度距离, 结合向量空间模型构建医疗术语信息元数据特征空间。在元数据特征空间内, 依据元数据的时间属性, 对各类数据进行合并处理, 以此同步数据索引, 并结合多级索引集群与上下文编码增强语义特征。引入基于一致性正则化与熵最小化原则的 MixMatch 半监督学习算法, 利用少量标注数据对无标签数据进行数据增强, 从而生成无标签数据的低熵伪标签, 增强语义特征表达, 并计算检索词与候选信息的相似度, 实现跨类别或近义术语的高效匹配。实验结果表明, 应用设计的方法进行医疗术语信息关键词检索, 输出的检索结果与输入关键词适配度高于 95%, 可以提高信息检索的精度。

Abstract

Medical terminology is characterized by strong heterogeneity and scarce labeled data, leading to weak semantic feature representation and low retrieval accuracy. To address these issues, a semi-supervised learning-based fuzzy retrieval algorithm for medical term information keywords is proposed. Word embeddings are utilized to standardize and map original medical term information. Based on core metadata information nodes and user query information nodes, the confidence distance between retrieval samples is determined, and a vector space model is employed to construct a metadata feature space for medical term information. Within this metadata feature space, data are merged according to their temporal attributes, thereby synchronizing data indexes. Semantic features are enhanced through multi-level index clustering and contextual encoding. The MixMatch semi-supervised learning algorithm, based on consistency regularization and entropy minimization principles, is introduced. A small amount of labeled data is used for data augmentation of unlabeled data, generating low-entropy pseudo-labels for unlabeled data to enhance semantic feature representation. The similarity between retrieval terms and candidate information is then calculated, enabling efficient matching across categories or synonymous terms. Experimental results demonstrate that applying the proposed method for medical term information keyword retrieval yields a retrieval result relevance exceeding 95%, effectively improving information retrieval accuracy.

关键词

半监督学习 / 医疗术语 / 路径 / 信息关键词 / 模糊检索

Key words

semi-supervised learning / medical terminology / path / information keywords / fuzzy retrieval

参考文献

[1] 唐秀, 伍赛, 侯捷, 等. 面向多模态模型训练的高效样本检索技术[J]. 软件学报, 2024, 35(3): 1125-1139.

[2] TANG X, WU S, HOU J, et al. Efficient Sample Retrieval Techniques for Multimodal Model Training[J]. Journal of Software, 2024, 35(3): 1125-1139.

[3] 李岩, 张敏艺, 宿汉辰, 等. 基于跨模态相似度学习的端到端不规则文本检索方法[J]. 无线电工程, 2023, 53(3): 501-507.

[4] LI Y, ZHANG M Y, SU H C, et al. End-to-End Irregular Text Retrieval via Cross-Model Similarity Learning[J]. Radio Engineering, 2023, 53(3): 501-507.

[5] 师广田, 张峰, 张辉, 等. 标签驱动语义感知学习的跨模态哈希检索方法[J]. 重庆邮电大学学报(自然科学版), 2025, 37(6): 870-883.

[6] SHI G T, ZHANG F, ZHANG H, et al. Label-Driven Semantic-Aware Learning for Cross-Modal Hashing Retrieval[J]. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition), 2025, 37(6): 870-883.

[7] 贾义, 毛佳昕. 基于弱监督学习的稠密向量检索模型[J]. 中文信息学报, 2025, 39(10): 156-166.

[8] JIA Y, MAO J X. Dense Retrieval Models Based on Weakly Supervised Learning[J]. Journal of Chinese Information Processing, 2025, 39(10): 156-166.

[9] 李骜, 谢委衡, 高天宇, 等. 基于多标签语义相似性引导的跨模态哈希检索方法[J]. 光电子·激光, 2025, 36(10): 1063-1077.

[10] LI A, XIE W H, GAO T Y, et al. Cross-Modal Hashing Retrieval Method Based on Multi-Label Semantic Similarity Guiding[J]. Journal of Optoelectronics·Laser, 2025, 36(10): 1063-1077.

[11] 肖泉彬, 陈源, 吴毅坚, 等. 基于代码克隆差异分析的函数模板挖掘和检索方法[J]. 软件学报, 2025, 36(6): 2774-2793.

[12] XIAO Q B, CHEN Y, WU Y J, et al. Function Template Mining and Retrieval Based on Code Clone Difference Analysis[J]. Journal of Software, 2025, 36(6): 2774-2793.

[13] 彭姣, 贺月, 商笑然, 等. 基于渐进原型匹配的文本-动态图片跨模态检索算法[J]. 计算机科学, 2025, 52(9): 276-281.

[14] PENG J, HE Y, SHANG X R, et al. Text-Dynamic Image Cross-Modal Retrieval Algorithm Based on Progressive Prototype Matching[J]. Computer Science, 2025, 52(9): 276-281.

[15] 王丹, 张峰, 张辉, 等. 基于信息互补与交叉注意力的跨模态检索方法[J]. 计算机应用研究, 2025, 42(7): 2032-2038.

[16] WANG D, ZHANG F, ZHANG H, et al. Information Complementarity and Cross-Attention for Cross-Modal Retrieval[J]. Application Research of Computers, 2025, 42(7): 2032-2038.

[17] 陈曦, 彭姣, 张鹏飞, 等. 基于预训练模型和编码器的图文跨模态检索算法[J]. 北京邮电大学学报, 2023, 46(5): 112-117.

[18] CHEN X, PENG J, ZHANG P F, et al. Cross-Modal Retrieval Algorithm for Image and Text Based on Pre-Trained Models and Encoders[J]. Journal of Beijing University of Posts and Telecommunications, 2023, 46(5): 112-117.

[19] 夏鑫雨, 朱磊, 聂秀山, 等. 典型概念驱动的模态缺失深度跨模态检索[J]. 计算机辅助设计与图形学学报, 2025, 37(3): 519-532.

[20] XIA X Y, ZHU L, NIE X S, et al. Typical Concept-Driven Modality-Missing Deep Cross-Modal Retrieval[J]. Journal of Computer-Aided Design & Computer Graphics, 2025, 37(3): 519-532.

[21] 叶亚芬, 原德巍. 基于数据化和文本检索技术的档案资源智能聚类研究[J]. 浙江档案, 2023(8): 17-19.

[22] YE Y F, YUAN D W. Research on Intelligent Clustering of Archival Resources Based on Data Digitization and Text Retrieval Technology[J]. Zhejiang Archives, 2023(8): 17-19.

[23] 杨梦雅, 赵琰, 薛亮. 基于改进的 Vision Transformer 深度哈希图像检索[J]. 陕西科技大学学报, 2025, 43(4): 183-191.

[24] YANG M Y, ZHAO Y, XUE L. Deep Hashing Method Based on Improved Vision Transformer[J]. Journal of Shaanxi University of Science & Technology, 2025, 43(4): 183-191.

[25] 齐丹丹, 王长征, 郭少茹, 等. 基于主题多视图表示的零样本实体检索方法[J]. 广西师范大学学报(自然科学版), 2025, 43(3): 23-34.

[26] QI D D, WANG C Z, GUO S R, et al. Topic-Based Multi-View Entity Representation for Zero-Shot Entity Retrieval[J]. Journal of Guangxi Normal University (Natural Science Edition), 2025, 43(3): 23-34.

[27] 肖玥. 跨学科知识生成导向的交互式信息检索模型构建[J]. 图书馆建设, 2025(2): 143-153.

[28] XIAO Y. Construction of Interactive Information Retrieval Model Guided by Interdisciplinary Knowledge Generation[J]. Library Development, 2025(2): 143-153.

[29] 黄溪, 王先兵, 林海, 等. 基于自集成视觉 Transformer 的图像检索[J]. 武汉大学学报(工学版), 2024, 57(12): 1795-1802.

[30] HUANG X, WANG X B, LIN H, et al. Image Retrieval Based on Self-Ensemble Vision Transformer[J]. Engineering Journal of Wuhan University, 2024, 57(12): 1795-1802.

[31] 唐莹莹, 陈玉玲, 罗运, 等. 基于全同态加密的可验证多关键词密文检索方案[J]. 计算机工程, 2025, 51(4): 188-197.

[32] TANG Y Y, CHEN Y L, LUO Y, et al. Verifiable Multi-Keyword Ciphertext Retrieval Scheme Based on Fully Homomorphic Encryption[J]. Computer Engineering, 2025, 51(4): 188-197.

[33] 赵周颖, 余正涛, 黄于欣, 等. 基于要素关联图的汉越跨语言事件检索方法[J]. 现代电子技术, 2024, 47(7): 127-132.

[34] ZHAO Z Y, YU Z T, HUANG Y X, et al. Chinese-Vietnamese Cross-Lingual Event Retrieval Method Based on Arguments Relational Graph[J]. Modern Electronics Technique, 2024, 47(7): 127-132.

基金资助

江苏省社科应用研究精品工程重点基金资助项目(25SYA-025)