地理大数据挖掘研究进展与挑战

大数据时代,全面涵盖人类活动与地理环境信息的地理大数据为更全面认识“人-地”关系提供了新的机遇。数据挖掘是地理大数据产生“大价值”的关键。与传统目的性采样数据(或“小数据”)相比,地理大数据具有更细的时空粒度、更广的时空范围、更丰富的人地关系信息、更高的时空有偏性及更低的时空精度。地理大数据的独特性使得地理大数据挖掘面临新的挑战。本文首先对地理大数据挖掘与空间数据挖掘的区别与联系进行分析;然后,对当前地理大数据挖掘方法、应用及软件的研究进展进行回顾和总结;最后,对地理大数据挖掘面临的挑战和发展趋势进行了展望。通过对地理大数据挖掘研究进展进行系统的分析,有望为地理大数据挖掘理论与方法的完善提供一定的参考和借鉴。

Recent advance and challenge in geospatial big data mining

In the era of big data, geospatial big data provide new opportunities for understanding complex human-land relationships. Data mining is essential for revealing valuable spatio-temporal patterns (e.g., clusters, outliers, association rules, etc.) hidden in geospatial big data. Geospatial big data has some unique characteristics, e.g., fine spatio-temporal granularity, wide spatio-temporal scope, rich information on human-land relationships, high spatio-temporal bias, and low spatio-temporal precision. Geospatial big data requires specially designed data mining methods given its unique characteristics. In this study, we first analyzed the relationships between spatial data mining and geospatial big data mining, then reviewed recent advances in geospatial big data mining, and finally summarized the challenges and further research directions of geospatial big data mining. This review is expected to provide some valuable reference for the improvement of theories and methods of geospatial big data mining.