Addiction Nexus – RDF Knowledge Graph (TBox + ABox)

This dataset provides the updated Addiction Nexus Knowledge Graph (ADNXKG) in RDF/Turtle (.ttl) format, encompassing both the TBox (ontology/schema) and ABox (instance data). Building on the previous version, this v2.0 release includes 2,305,987 RDF triples, an increase from the initial dataset, and integrates 50,000 records of addiction determinants sourced from Zenodo. The graph is QB4OLAP-compliant, modeling hierarchical dimensions (Country, Lifestyle, Mental Health, Occupation, Gender) and measures (e.g., Age, Stress Levels, Smoker Counts, Alcohol Consumption) to support multidimensional OLAP analytics. New features include enhanced socioeconomic correlations (e.g., 788 smokers in low-education cohorts) and OWL connectivity validated via SPARQL (e.g., owl:sameAs links to Wikidata). The dataset is designed for OLAP operations (roll-up, drill-down, slice, dice), knowledge graph exploration, and FAIR-compliant reuse, achieving 91.67% FAIR compliance as detailed in the associated paper [DOI to be inserted]. Key updates include refined TBox semantics, additional ABox instances reflecting recent analysis (August 2025), and alignment with the paper’s findings on performance and scalability (sub-second response times across 10,000–200,000 triples). The file is intended for researchers in addiction analytics, semantic web technologies, and data-driven health interventions.本数据集提供了更新版的成瘾关联知识图谱（Addiction Nexus Knowledge Graph, ADNXKG），采用RDF/Turtle（.ttl）格式存储，涵盖TBox（本体/模式层）与ABox（实例数据层）两部分。本版本基于既往数据集迭代而来，作为2.0版发布，共包含2,305,987条资源描述框架（RDF）三元组，相较于初始数据集规模有所提升，同时整合了来自Zenodo的50,000条成瘾影响因素记录。本知识图谱符合QB4OLAP标准，构建了分层维度（国家、生活方式、心理健康、职业、性别）与度量指标（如年龄、压力水平、吸烟者数量、酒精摄入量），以支持多维在线分析处理（Online Analytical Processing, OLAP）分析。新增特性包括增强的社会经济相关性（如低教育水平群体中的788名吸烟者），以及通过SPARQL（SPARQL RDF查询语言）验证的Web本体语言（OWL）连通性，例如指向Wikidata的owl:sameAs链接。 本数据集专为OLAP操作（上卷、下钻、切片、切块）、知识图谱探索以及符合FAIR（可发现、可访问、可互操作、可重用）标准的复用而设计，合规率达91.67%，相关细节已在附带论文[DOI待补充]中详述。核心更新内容包括优化的TBox语义、新增的ABox实例（反映2025年8月的最新分析结果），以及与论文中关于性能与可扩展性的研究结论保持一致（在10,000至200,000条三元组的规模下，响应时间可控制在亚秒级）。本数据集面向成瘾数据分析、语义网技术以及数据驱动型健康干预领域的研究人员。