基于微博大数据的西安国内游客日内时间分布模式研究

A STUDY ON THE TEMPORAL DISTRIBUTION PATTERN OF DOMESTIC TOURISTS IN XI'AN BY HOURS BASED ON THE MICROBLOG BIG DATA

移动互联网的迅速发展与大数据时代的到来为研究不同尺度的游客时间行为提供了可能。本文以西安为案例地,基于游客生成的大规模微博签到数据研究国内游客的日内时间分布模式。研究表明:1国内游客与日常用户的分时签到规律存在明显的差异;2游客在重要旅游节点的日内时间分布模式可分为"多峰波动型"、"日间活跃型"和"夜间活跃型";3客流的时间分布模式与空间特征具有较强的对应关系;4空间与活动的叠加是形成游客时间分布模式的重要作用机制。

The researches on temporal behaviors rules of tourist flow are usually made from macro perspective rather than micro perspective. With the advent of the mobile Internet and big data, the researches on temporal behavior rules of tourist flow is entering a new stage. With the example of Xi＇an, this paper studies the temporal distribution pattern of domestic tourists by hours based on the microblog big data which are generated by tourists. 388,676 geotagged messages were obtained through calling the open API of Sina Microblog. After filtering, 131,960 available data are reserved for studying. First, from the general perspective, this paper studies the comparison of check-in pattern of domestic tourists by hours in Xi＇an and ordinary users,then explores the temporal distribution pattern of tourists by hours in important travel notes. The study found that： 1there are greater differences in check-in pattern by hours between domestic tourists in Xi＇an and ordinary users, including the entire shape of check-in pattern, the check-in frequency in day and night, the delay and hysteresis of tourists behavior. 2the temporal distribution patterns of tourist flow by hours in important travel notes can be divided into three types： multi-peak fluctuating pattern, daytime active pattern and nighttime active pattern. The multi-peak fluctuating pattern mainly includes transportation hubs, which shows several peaks and flexible temporal distribution in temporal distribution figure. 3the temporal distribution patterns of domestic tourists by hours have corresponding relationships with spatial characteristics. 4the impact of space and activity on temporal distribution patterns is an important mechanism.