本文選題：地理位置簽到 + 空間相互作用　； 參考：《北京大學(xué)學(xué)報(自然科學(xué)版)》2017年05期

【摘要】：基于社交網(wǎng)絡(luò)大數(shù)據(jù)的研究視角,選取全國348個城市之間一年的跨城市社交媒體地理位置簽到數(shù)據(jù),采取優(yōu)化粒子群(PSO)方法,使用引力模型,逆向推導(dǎo)該系統(tǒng)中空間相互作用的距離衰減函數(shù)以及各城市的節(jié)點(diǎn)吸引力。通過引入經(jīng)濟(jì)發(fā)展水平、產(chǎn)業(yè)結(jié)構(gòu)、人口規(guī)模和結(jié)構(gòu)、旅游競爭力、教育水平5個方面的12項變量,經(jīng)過因子分析和回歸分析,探究這些變量對節(jié)點(diǎn)吸引力的影響作用。結(jié)果表明,社交媒體簽到系統(tǒng)中的交互流量符合距離衰減的冪律函數(shù),與其他交互系統(tǒng)相比,其距離衰減系數(shù)偏小,說明全國尺度的城市間人的移動受距離影響不明顯。對全國城市節(jié)點(diǎn)吸引力及其排名的進(jìn)一步分析發(fā)現(xiàn),與旅游競爭力、城市發(fā)展成熟度、人口規(guī)模這幾個維度相關(guān)的因子對社交媒體簽到系統(tǒng)中的城市節(jié)點(diǎn)吸引力有顯著的影響。研究結(jié)論將為更好地理解人類簽到和移動行為,為進(jìn)一步了解復(fù)雜網(wǎng)絡(luò)系統(tǒng)中節(jié)點(diǎn)吸引力的內(nèi)涵做出一定的理論和實(shí)際貢獻(xiàn)。
[Abstract]:Based on the research perspective of social network large data, this paper selects the data of cross city social media geographic location between 348 cities of the country for one year, adopts the optimization particle swarm (PSO) method, uses the gravitational model, and deduces the distance attenuation function of the spatial interaction in the system and the attraction of the nodes of each city. There are 12 variables in 5 aspects: exhibition level, industrial structure, population size and structure, tourism competitiveness and education level. Through factor analysis and regression analysis, the influence of these variables on the attraction of nodes is explored. The results show that the interaction flow in the social media signature system meets the power law function of distance attenuation and is associated with other interactive systems. The distance attenuation coefficient is small, which indicates that the distance between urban people is not affected by the distance of the national scale. Further analysis of the attraction and ranking of the national urban nodes finds that the factors related to tourism competitiveness, urban development maturity, and population size are related to the urban nodes in the social media sign system. Gravity has a significant impact. The research conclusions will be a better understanding of human signature and mobile behavior, and make some theoretical and practical contributions to further understanding the connotation of node attraction in complex network systems.
【作者單位】： 北京大學(xué)城市與環(huán)境學(xué)院旅游研究與規(guī)劃中心;北京大學(xué)遙感與地理信息系統(tǒng)研究所;
【基金】：國家自然科學(xué)基金(41271151)資助
【分類號】：F299.23;TP18
，

本文編號：1920593