GeoGLUE

该数据集是一个评估地理语言理解能力的基准，它包含了六个自然语言理解任务。这些任务涵盖了地理自然语言处理的多个方面，以大规模收集自公开发布的地理资源的数据为支撑。具体任务包括地理文本相似性评估、地理元素标注、地理组成分析、地理情境分割以及地理实体对齐等。

This dataset is a benchmark for evaluating geographical language understanding capabilities. It comprises six natural language understanding tasks. These tasks cover multiple aspects of geographical natural language processing, and are supported by large-scale data collected from publicly released geographical resources. The specific tasks include geographical text similarity evaluation, geographical element annotation, geographical composition analysis, geographical context segmentation, and geographical entity alignment, among others.