透過您的圖書館登入
IP:3.21.127.68
  • 學位論文

應用推特資料探討使用者差異的都市公園服務範圍: 以東京都立公園為例

Using Twitter Data for Identifying Urban Park Service Areas among Different Users: a Case Study of Tokyo Parks

指導教授 : 溫在弘

摘要


隨著都市化程度提高與人口過度集中都市,為提升都市生活品質並增進都市居民的休閒遊憩機會,公園的興建、整治是都市規劃的重點。為此,政府建立公園分類制度,以期透過合理評估將服務資源進行分配。過去多以行政區居民人口、公園面積進行公共設施服務範圍評估,卻面臨可調整性面積單元問題 (MAUP)。後基於地理資訊系統使用,實際地理因素與變異如道路距離、車速差異得以被考量,然而,公園在週間或假日、基於當地使用或賞櫻、賞楓等旅遊目的的實際使用情況卻難以進行討論。而隨著行動裝置與GPS日益普及,高精度、豐富個體時空資訊得以被蒐集儲存,其中,社群平台具有時間即時、取得便利性與資料細緻等優勢,並可藉由演算方法反映發佈者屬性和特性,透過社群媒體資料使用,將得以分析公園使用情況並探討其實際服務範圍。 本研究以東京作為研究區域,自推特收集東京都18處綜合公園50,303推文資料,藉由時窗進行使用者判別,結果顯示51.4% 為國內外旅客造訪,接著,透過Twitter-DBSCAN取得都市居民居住地,並以國內、外國旅客旅遊住宿視為其臨時居住地,發現各使用者數量皆隨距離呈現遞減現象;同時,以第三四分位數作為最大服務範圍標準,發現外國旅客由於交通方式限制集中都心商業區,實際服務範圍最小,國內旅客具備自有小客車、移動範圍較大實際服務範圍最大,都市居民則介於兩者之間。最後,進一步透過東京都土地利用分類、週間週末、季節分別對於使用者居住地分佈熱點進行探討:偏遠獨立住宅區綜合公園吸引東京都心商業區都市居民使用;都市居民存在週間、週末公園使用偏好差異,國內外旅客則反之;在季節差異上,不同使用者並未存在居住地分佈的明顯差異。

並列摘要


Urban parks play an essential role in the city as public places. Therefore, governments set up with park classification system by evaluating service area, in order to supply park services to city residents equally, but however, the actual use of the urban park is difficult to discuss. Instant time, convenience and detailed information as the advantages of social media, reflecting the users' attributes and characteristics. For this reason, with the use of social media materials, it will be possible to analysis the differences of urban park service area among users. In this research, first we take Tokyo as our research area, collecting 50,303 tweets of 18 parks in Tokyo, and a total of 2.7 million geotag tweets of park users. Second, the time window grouping result shows that 51.4% are visited by domestic and foreign tourists. After classifying users and finding the corresponding inhabitation of residents, it is found that the number of residents decreases with distance. In addition, foreign tourists concentrate on the area of the center due to restriction of transportation means, so the actual service area is the smallest. Domestic tourists may have their own cars, therefore the actual service area is the largest. Then the urban residents are somewhere in between above. Lastly, we further explore the distribution hotspots of users' residences through the land use classification, weekdays and seasons in Tokyo. First we finds parks in remote residential areas attract urban residents in the Tokyo metropolitan business district.Second, urban residents have differences in the use of parks during the week and weekends, while domestic and foreign passengers do not. Third, In terms of seasonal differences, there is no significant difference in the distribution of residences.

並列關鍵字

Twitter urban park crowd classification service area DBSCAN

參考文獻


張曜麟、洪于婷、謝琦強、蔣文樺(民101)。都市公園體驗認知與服務範圍之探討。嘉南學報,38,415-428。
賴哲三(民65)。都市公園綠地計畫。台北:捷盛書局。
Aday, L. A., & Andersen, R. (1974). A framework for the study of access to medical Care, Health Services Research, 9(3): 208-220.
Chen, W., & Wang, Y. F. (2009). An Equity Evaluation of Urban Park Location-Allocation——A Case Study Within the Outer Ring of Shanghai City. Journal of Anhui Normal University (Natural Science), 32(4): 373-377.
Chen, Z., Hua, C., Liu, G., Liu, Q., Li, H., Wang, Q., & Li, X. (2017). Analysis of Spatial-temporal Characteristics of Resident Travel for Hospitals Based on Taxi GPS Data. Journal of Geo-information Science, 20(8): 1111-1122.

延伸閱讀