Abstract: As a bridge connecting data, knowledge, and intelligence, knowledge graph has been widely applied in such as search assistance, intelligent recommendation, question-answering systems, and natural language processing. However, with the expansion of application scenarios, static knowledge graph gradually has shown limitations in handling dynamic knowledge. The emergence of temporal knowledge graph addresses this shortcoming by integrating temporal information into the graph structure, enabling a more accurate representation of the dynamic changes in knowledge. This paper provides a comprehensive study on the construction of temporal knowledge graph. It begins by introducing the concept of temporal knowledge graph and clarifying its value in handling dynamic knowledge. It then delves into the construction process of temporal knowledge graph, dividing the core process into three key stages: knowledge extraction, knowledge fusion, and knowledge computing. Subsequently, we have thoroughly organized each stage, each stage is detailed with task definitions, research summaries, and the application of large language models. In the knowledge extraction stage, we focus on named entity recognition, relation extraction, and time information extraction; in the fusion stage, entity alignment and entity linking are discussed; and in the computation phase, we emphasis on knowledge reasoning. Finally, we explore the challenges faced at each stage and look forward to future research directions, providing insights for enhancing temporal knowledge graph quality.

Key words: temporal knowledge graph, knowledge extraction, temporal information extraction, knowledge fusion, knowledge reasoning

摘要： 知识图谱作为连接数据、知识和智能的桥梁，已被广泛应用于辅助搜索、智能推荐、问答系统、自然语言处理等多个领域。然而，随着应用场景的不断拓展，传统静态知识图谱逐渐暴露出在处理动态知识方面的局限性。时序知识图谱的出现弥补了这一缺陷，它将时间信息融入图谱结构，能够更准确地表示知识的动态变化。对时序知识图谱的构建进行了全面的研究，首先介绍了时序知识图谱的概念，明确了其在处理动态知识时的价值。接着，解析了时序知识图谱构建流程，将其核心过程划分为知识抽取、知识融合和知识计算三大环节。然后，对每个阶段进行了梳理，明确了任务定义、总结了研究现状，并探讨了大语言模型在这些任务中的应用。在知识抽取阶段，重点关注命名实体识别、关系抽取和时间信息抽取；在知识融合阶段，探讨了实体对齐和实体链接；在知识计算阶段，聚焦于知识推理。最后，深入分析了每个阶段面临的挑战，并针对特有挑战展望了未来的研究方向，为时序知识图谱质量的提升提供借鉴和启示。

关键词: 时序知识图谱, 知识抽取, 时间信息抽取, 知识融合, 知识推理