發(fā)布時間：2018-03-23 12:26

  本文選題：地理空間數(shù)據(jù)　切入點：社交媒體　出處：《地理科學進展》2017年09期

【摘要】："城市意象"研究對城市文化感知、城市管理與規(guī)劃、旅游資源開發(fā)等具有重要意義。近年來,隨著智能移動終端和社交媒體的普及,產生了大量城市內包含有文本和地理位置等信息的社交媒體數(shù)據(jù),涉及城市的各個區(qū)域,為開展城市意象的綜合感知研究提供了新的途徑。本文以2016年北京市帶位置簽到的新浪微博數(shù)據(jù)為例,在空間聚類發(fā)現(xiàn)熱點區(qū)域的基礎上,采用詞頻—逆文件頻率(TF-IDF)與文檔主題生成模型LDA兩類典型的文本分析的方法,挖掘城市不同熱點區(qū)域的主題,以感知北京市不同熱點區(qū)域的社會文化功能和人群行為,并在此基礎上通過對熱點區(qū)域高頻主題詞進行共詞聚類分析,深度挖掘北京市的總體意象。研究表明,運用文本挖掘及地理大數(shù)據(jù)分析的城市意象研究方法,能及時感知人群在城市不同場所的活動、態(tài)度、偏好,從而揭示城市的社會文化及功能特征,是對刻畫城市物質形態(tài)的城市意象五要素模型的重要補充。此外,以北京市熱點區(qū)域為例的實證研究結果對現(xiàn)實中的城市特色傳承與空間品質塑造等有一定的啟發(fā)意義。
[Abstract]:The study of "urban image" is of great significance to urban cultural perception, urban management and planning, tourism resource development, etc. In recent years, with the popularity of intelligent mobile terminals and social media, A large number of social media data in cities, including text and geographical location, have been generated, covering all parts of the city, This paper takes the data of Sina Weibo with location check-in in 2016 as an example, based on the spatial clustering to find hot spots, Two typical text analysis methods, word frequency inverse document frequency (TF-IDF) and document theme generation model (LDA), are used to excavate the themes of different hot spots in cities in order to perceive the social and cultural functions and crowd behavior of different hot spots in Beijing. On the basis of this, this paper makes a coterm clustering analysis of high-frequency thematic words in hot spots and deeply excavates the overall image of Beijing. The research shows that the research method of urban image is based on text mining and geographical big data analysis. The ability to perceive the activities, attitudes and preferences of people in different places in a city in time is an important supplement to the five-element model of city image, which can reveal the social, cultural and functional characteristics of the city. The results of the empirical research on the hot spot areas in Beijing are instructive to the inheritance of urban characteristics and the shaping of spatial quality in reality.
【作者單位】： 北京大學遙感與地理信息系統(tǒng)研究所;北京聯(lián)合大學旅游信息化協(xié)同創(chuàng)新中心;
【基金】：國家自然科學基金項目(41501162) 北京市社會科學基金項目(17JDGLB002) 北京聯(lián)合大學人才強校優(yōu)選計劃(BPHR2017DS08)~~
【分類號】：K901.6
，

本文編號：1653447