主流价值观知识数据
收藏资源简介：
主流价值观知识数据主要面向主流价值观传播与认知研究、精准传播建模需求建设，其核心目标是构建覆盖全面、来源可信、内容多样的主流价值观语料库，为主流价值观知识体系的构建与研究提供坚实的数据基础。该数据集基于人民日报、环球时报等权威媒体平台的新闻数据产生，整合了2012年至2024年的新闻报道，主要记录了时政、社会热点、民生新闻以及负面信息等数据，旨在为构建主流价值观理论知识体系提供高质量的数据支撑，助力相关领域的深入研究与应用开发。 主流价值观知识数据，数据量约30,500条，数据大小约245M左右。数据文件knowledge.json包含知识条目和文档信息，以支持多样化的价值观分析任务。知识条目部分分为四个类别：一级知识条目（名称first-level）、二级知识条目（名称second-level）、三级知识条目（名称third-level）和负面知识条目（名称negative）。其中，一级知识条目直接来源于权威性最高的内容，如人民日报社和党史文献，代表主流价值观的核心内容；二级和三级知识条目则逐级扩展知识体系的覆盖范围，涵盖次核心和外围知识，分别从官方新闻到行业媒体等来源提取。负面知识条目从境外媒体中抽取，主要记录与主流价值观相悖的信息，以便在舆情分析和矛盾对比研究中提供支持。每条知识条目还附带节点ID、实体类别、情感倾向、来源列表和相关节点等属性，能够实现精细化分析。
Mainstream Values Knowledge Dataset is primarily developed to meet the needs of mainstream values communication and cognition research, as well as precise communication modeling. Its core objective is to build a comprehensive, credibly-sourced, and diverse corpus of mainstream values, providing a solid data foundation for the construction and research of the mainstream values knowledge system. This dataset is generated from news data of authoritative media platforms including People's Daily and Global Times, integrating news reports from 2012 to 2024. It mainly records data such as current politics, social hotspots, livelihood news, and negative information, aiming to provide high-quality data support for the construction of the theoretical knowledge system of mainstream values, and facilitate in-depth research and application development in related fields. The Mainstream Values Knowledge Dataset contains approximately 30,500 entries, with a total size of around 245 MB. The data file `knowledge.json` stores knowledge entries and document information to support diverse values analysis tasks. The knowledge entries are divided into four categories: first-level knowledge entries (named `first-level`), second-level knowledge entries (named `second-level`), third-level knowledge entries (named `third-level`), and negative knowledge entries (named `negative`). Among them, first-level knowledge entries are directly sourced from the most authoritative content, such as content from People's Daily and Party history documents, representing the core content of mainstream values. Second-level and third-level knowledge entries gradually expand the coverage of the knowledge system, covering sub-core and peripheral knowledge, and are extracted from sources ranging from official news to industry media respectively. Negative knowledge entries are extracted from overseas media, mainly recording information that conflicts with mainstream values, to provide support for public opinion analysis and contradiction comparison research. Each knowledge entry also includes attributes such as node ID, entity category, sentiment tendency, source list, and related nodes, enabling refined analysis.
