随着工程建设行业信息化和数字化程度逐步加强，工程建设过程中所产生的相关数据量呈现逐渐增加的趋势。然而，传统的建设管理模式以人工为主， 存在着信息覆盖面狭窄、信息交互不方便、过程管理耗时费力等突出问题，这凸显了建立信息化工程建设质量管理手段的必要性和紧迫性。 质量管控知识库作为工程质量管理信息化工作开展的基石和实现信息平台功能的底层基础，包含了更加精细化、组织化和结构化的质量管控信息，能够适应信息化工作开展的需要，为实现施工质量的数字化和智能化管控提供基础。
本文针对我国建筑行业施工质量管控及其信息化现状，基于 POP 模型及施工产品—过程模型构建了施工质量管控知识库，主要开展了以下工作。
首先， 基于 POP 模型， 分析了从产品（P）、组织（O）、过程（P） 三个维度确定施工质量管控知识， 研究了从法律法规、规范标准和其他途径（如施工经验、工程案例等） 三个来源获取质量管控知识的方法， 保证了质量管控知识的全面性和完整性。
其次，基于施工产品—过程模型设计了施工质量管控知识库的构架。 将工程项目分解为一系列的施工产品（对象），根据施工产品的施工建造特点划分其施工过程，设计了可涵盖质量管控内容“三个维度”和“三个来源”的质量管控知识库的构架。同时，基于知识图谱 neo4j 构建了图知识库。
第三，根据质量管控知识的“三个来源”中质量管控知识的不同特点，研究了不同的提取方式。对半结构化信息如规范标准条文等采用基于文本的方法自动抽取，对政策法规、其他途径（如施工经验、工程案例等）非结构化信息，采用人工标注方式提取。
最后，以某大型悬索桥梁的桩基工程项目为例，以表格形式直观展示了知识库输出的具体质量管控内容，并介绍了知识库在工程中的具体应用。

关键词：施工质量；标准化；POP 模型；知识库；质量管控知识

With the increasing degree of informatization and digitization in the engineering construction industry, there has been a gradual rise in the volume of relevant data generated during the construction process. However, the traditional manual construction management mode presents challenges such as limited information coverage, inconvenient information interaction, and time-consuming process management. This underscores the importance and urgency of establishing informatized means for engineering construction quality management. The quality control knowledge base is the foundation of engineering quality management informatization and serves as the basis for the functionality of an information platform encapsulating refined, organized, and structured quality control information. This knowledge base is designed to meet the requirements of informatization and facilitate the transition towards digital and intelligent construction quality control.
In this paper, we focus on the current state of construction quality control and informatization in China's construction industry, and construct a construction quality control knowledge base based on the POP model and the construction product-process model. The following key tasks are undertaken:
First, leveraging the POP model, we analyze the product, organization, and process dimensions to identify construction quality control knowledge and explore methods for acquiring such knowledge from sources including laws, regulations, norms, construction experience, and engineering cases. This approach ensures the comprehensiveness and integrity of quality control knowledge.
Next, the structure of the construction quality control knowledge base is designed based on the construction product-process model. Projects are decomposed into a series of construction products (objects), and the construction process is categorized based on the characteristics of these products. The knowledge base structure encompasses the three dimensions and sources of quality control content and employs a graph knowledge base built on the neo4j knowledge graph.
Furthermore, we study different extraction methods for quality control knowledge sourced from the "three dimensions." While structured information such as norms and standards is automatically extracted utilizing text-based methodologies, unstructured data from policies, regulations, and other sources are extracted through manual labeling.
Finally, using the pile foundation project of a large suspension bridge girder as a case study, the specific quality control content outputted from the knowledge base is visualized in a tabular format, and the application of the knowledge base in the project is demonstrated.