zilliz/wikipedia_zh_cn_500k-context-relevance-with-think

该数据集用于训练`zilliz/semantic-highlight-bilingual-v1`模型，用于在RAG（检索增强生成）系统中进行语义高亮。数据集包含带有上下文跨度相关性注释的查询-上下文对，帮助识别文档中与查询语义相关的部分，即使不包含精确关键词匹配。关键特征包括上下文跨度位置（`context_spans`）、相关性标签（`context_spans_relevance`）和思考过程（`think_process`），后者保留了标注过程中的推理过程以提高准确性和可解释性。数据集来源于`fjcanyue/wikipedia-zh-cn`，遵循GNU自由文档许可证1.3和知识共享署名-相同方式共享4.0国际许可协议。

This dataset is used for training the `zilliz/semantic-highlight-bilingual-v1` model for semantic highlighting in RAG (Retrieval-Augmented Generation) systems. It contains query-context pairs with relevance annotations for context spans, helping to identify which parts of a document are semantically relevant to a query, even without exact keyword matches. Key features include context span positions (`context_spans`), relevance labels (`context_spans_relevance`), and the think process (`think_process`), which preserves the reasoning process during annotation to improve accuracy and interpretability. The dataset is derived from `fjcanyue/wikipedia-zh-cn` and follows the GNU Free Documentation License 1.3 and Creative Commons Attribution-ShareAlike 4.0 International license.