唐炉亮, 邹倩倩, 张霞, 任畅, 李清泉. 融合线上线下轨迹的景观热度评价[J]. 武汉大学学报 ( 信息科学版), 2018, 43(11): 1704-1711. DOI: 10.13203/j.whugis20170043
引用本文: 唐炉亮, 邹倩倩, 张霞, 任畅, 李清泉. 融合线上线下轨迹的景观热度评价[J]. 武汉大学学报 ( 信息科学版), 2018, 43(11): 1704-1711. DOI: 10.13203/j.whugis20170043
TANG Luliang, ZOU Qianqian, ZHANG Xia, REN Chang, LI Qingquan. Online and Offline Trajectory Data Fusion and Evaluation of Landscape Attractions[J]. Geomatics and Information Science of Wuhan University, 2018, 43(11): 1704-1711. DOI: 10.13203/j.whugis20170043
Citation: TANG Luliang, ZOU Qianqian, ZHANG Xia, REN Chang, LI Qingquan. Online and Offline Trajectory Data Fusion and Evaluation of Landscape Attractions[J]. Geomatics and Information Science of Wuhan University, 2018, 43(11): 1704-1711. DOI: 10.13203/j.whugis20170043

融合线上线下轨迹的景观热度评价

Online and Offline Trajectory Data Fusion and Evaluation of Landscape Attractions

  • 摘要: 人们在线上网络空间和线下现实空间活动时, 会产生大量线上线下的痕迹(轨迹), 这种轨迹作为一种讯息会影响其他人的活动。现有研究大多单一地采用问卷调查、网络数据、GPS轨迹等方法, 忽略了移动互联时代线上线下空间活动的关联性以及历史活动足迹对日后活动的影响。基于对搜索量、签到和照片等网络大数据与线下GPS轨迹大数据的融合, 对人们线上线下活动进行时空行为模式分析, 建立基于信息素和回归分析的计算模型, 实现了对景观休憩空间的吸引力评价。以采样接近5 a的181个用户轨迹数据及相应时间段的网络大数据, 对北京28个景观进行实验, 并与单一线上或线下数据的研究方法进行比较, 结果表明所提方法不仅综合体现了景观热度, 且能够对未来的潜在游憩价值进行估计。

     

    Abstract: In real space and cyberspace, people leave both online and offline traces of their daily lives, which might affect the behavior of other people.This pattern is similar to insect foraging behaviors based on pheromone reaction.Previous landscape evaluation only used questionnaires, network data, or GPS data, ignoring the relationship between online and offline activities in this mobile internet era, as well as overlooking the influence of previous activities on future behaviors.In this paper, the new model, based on the idea of pheromone, couples big website data with GPS trajectory data to analyze spatio-temporal recreational behaviors in real space and cyberspace, and finally evaluates the attractiveness of locations.To test the method, twenty-eight landscapes in Beijing were evaluated, using big website data and GPS trajectory datasets collected from 181 users over a period of about five years.The results show that this new method could be very promising, which outperforms the methods that only consider online data or offline data.

     

/

返回文章
返回