SignData

中国手语知识图谱数据集 v1.0 描述 这是中国手语知识图谱数据集的初始版本,旨在支持手语识别、手语教学及相关无障碍技术领域的研究和应用。 主要特点 数据集内容: - 3种类型的实体:手语词汇、身体部位和动作行为 - 2种预定义的关系类型链接实体:"执行"和"表达" - 基于四元组结构:(手语词汇, 身体部位, 动作行为, 时序顺序) 四元组例子: "看, 右手食指, 分开, 1" "馄饨, 右手食指, 在左手掌心上一抹, 1" 包含的文件: - data.json: 原始训练对话数据集 - neo4j/neo4j-import/: Neo4j导入格式的CSV文件 - sign_words.csv: 手语词汇节点 - body_parts.csv: 身体部位节点 - actions.csv: 动作行为节点 - quadruples.csv: 完整四元组数据 - performs_relationships.csv: "执行"关系数据 - executes_relationships.csv: "表达"关系数据 - neo4j/scripts/: Neo4j导入和查询脚本 - Neo4j_Import_Guide.md: 数据集使用说明 祝您探索中国手语知识图谱愉快! Chinese Sign Language Knowledge Graph Dataset v1.0 Description This is the initial version of the Chinese Sign Language Knowledge Graph dataset, designed to support research and applications in sign language recognition, sign language education, and related accessibility technologies. Key Features of the Dataset: Entity Types: The dataset contains 3 types of entities: sign words, body parts, and action behaviors. Relationships: 2 predefined relationship types link entities ("PERFORMS" and "EXECUTES"). Quadruple Structure: (sign_word, body_part, action, temporal_order) Example of Quadruples: "看, 右手食指, 分开, 1" (Look, right index finger, separate, 1) "馄饨, 右手食指, 在左手掌心上一抹, 1" (Wonton, right index finger, wipe on left palm, 1) Included Files: - data.json: Original training conversation dataset - neo4j/neo4j-import/: CSV files in Neo4j import format - sign_words.csv: Sign word nodes - body_parts.csv: Body part nodes - actions.csv: Action behavior nodes - quadruples.csv: Complete quadruple data - performs_relationships.csv: "PERFORMS" relationship data - executes_relationships.csv: "EXECUTES" relationship data - neo4j/scripts/: Neo4j import and query scripts - Neo4j_Import_Guide.md: Dataset usage instructions Enjoy exploring the Chinese Sign Language Knowledge Graph dataset!