基于社交媒体数据的城市人群分类与活动特征分析

发布时间：2018-03-03 21:57

本文选题：位置签到数据　切入点：时间序列　出处：《地球信息科学学报》2017年09期 　论文类型：期刊论文

【摘要】：空间信息技术已开始进入全空间信息系统发展阶段,即将空间信息系统的范畴从传统测绘空间扩展到宇宙空间、室内空间、微观空间等可量测空间。位置大数据不仅是全空间信息系统的重要研究对象之一,而且也成为了广域全空间中了解人们生活方式以及城市动态变化的一种有效途径。本文基于社交媒体数据中的位置签到数据,提出一种不同于传统以社会经济属性为依据的城市人群分类方法。首先利用签到数据的时间序列构造矩阵模型;然后,通过分析用户签到活动的时间特征,采用K-means聚类算法和K近邻算法(K-NN)识别出具有不同时空行为特征的城市人群(静态居民、动态居民、通勤者以及访问者);最后,本文根据得到的人群分类结果,通过分析不同类型人群的时空间行为特征,发现不同类型人群时空间行为的差异性与潜在规律性,从而为表征城市人群的组成结构及特征,研究城市时空结构提供一种新的视角。
[Abstract]:Spatial information technology has begun to enter the development stage of full spatial information system, that is, the scope of spatial information system is extended from traditional mapping space to space space and indoor space. Location big data is not only one of the important research objects of the whole spatial information system, It has also become an effective way to understand the way people live and the dynamic changes of cities in a wide area. This paper is based on the location check-in data in social media data. This paper presents a method of urban population classification, which is different from traditional social and economic attributes. Firstly, a matrix model is constructed by using the time series of check-in data, and then, by analyzing the time characteristics of user check-in activities, K-means clustering algorithm and K-NNN algorithm are used to identify urban population (static residents, dynamic residents, commuters and visitors) with different spatial and temporal behaviors. By analyzing the characteristics of time-space behavior of different types of people, the differences and potential regularity of spatial behaviors of different types of people are found, so as to represent the composition and characteristics of urban population. The study of urban space-time structure provides a new perspective.
【作者单位】：电子科技大学资源与环境学院;电子科技大学大数据研究中心;
【基金】：国家重点研发计划资助项目(2016YFB0502300) 国家自然科学基金项目(41471332;41571392) 中央高校基本科研业务费专项资金资助(ZYGX2015J113)
【分类号】：C912.81;P208

【相似文献】