應用自發性地理資訊於國家公園之遊憩行為分析與遊客管理—以陽明山國家公園為例

近年來隨著國人對於休閒與生活品質的重視度提升，民眾對於休閒遊憩的需求與日俱增，而台灣地狹人稠，遊憩資源十分有限，因此產生遊憩使用集中的現象，國家公園成為民眾休閒旅遊重要的資源，然每年遊客量增加轉為環境壓力，故維持遊客良好遊憩體驗機會與遊憩品質，亦控制遊憩行為對環境的衝擊，是國家公園長久以來管理的關鍵課題。國家公園過去對遊客分布與移動行為的研究多用現地觀察與問卷、訪談方式，難精確掌握實際空間分布，成果大幅受到時間與空間的限制並耗費大量人力成本。近年定位系統、地理標籤等技術快速發展，大眾可自由上傳資訊至社群平台分享，此等自發式地理資訊(VGI)具取得便利、覆蓋區域全貌、資料細緻等優勢，反映發佈者行為特性，為空間行為研究帶來突破性契機。本研究應用社群平台中自發性地理資訊進行研究，希望幫助國家公園對遊憩行為之管理，因此主要目的有兩部分：了解遊憩行為、熱點分布與其環境現況評估遊憩行為的影響；了解國家公園整體呈現的意象與遊客的感知與態度。陽明山國家公園位處大臺北都會區，交通便捷且遊憩資源豐富，提供都市居民休閒選擇，一年可達1,900萬人次之龐大遊客量，故本研究選擇陽明山國家公園為研究案例，研究分為兩階段進行：第一階段，為指認國家公園內的遊憩熱點，利用Flickr社群平台抓取地理標籤資料進行密度群集分析，透過現地調查了解環境現況，並以時間資訊分析遊客的旅遊路徑，並以關連規則分析了解景點間的關聯性；第二階段，透過Instagram遊客上傳相片取代僱請遊客攝影法獲取影像資料，以內容分析法進行抽樣、建構類目及設定分析單元、資料的分析與推論，評估景點對大眾所呈現的意象，並透過Facebook文字評論分析，了解遊客對於意象的感受與態度。本研究第一階段用VGI鑑別出共14個遊憩熱點，依資料數量依序為陽明公園、竹子湖、冷水坑、大屯山步道、擎天崗等。這些熱點大致與官方歷年遊憩據點遊客量統計相符；但進一步檢視時間分布發現，春季遊客集中於竹子湖與陽明公園，秋季則集中於擎天崗；而平假日的遊客分布較過去調查數據分布平均；探究夜間資料，發現大屯山為重要夜間熱點。在現地調查結果發現，大多景點內遊客分布密集處多有良好的景觀，而這些觀景點大多亦有相應之休憩平台或是圍欄規範遊客行為，減少環境衝擊的發生，然而部分步道若寬度不足且未加以設施規範遊客行為，則遊客外擴至步道外的行為容易造成周邊植生的踐踏。此外，關聯規則分析結果顯示，陽明山國家公園之遊憩關聯景點位於西南側，包括大屯自然公園、大屯山、二子坪遊憩區、向天山、竹子湖、小油坑與陽明公園，關聯景點移動主要以步道做串連，顯示步道管理之重要性。第二階段分析整體國家公園在遊客照片中所呈現的意象，其最主要的意象元素為動植物，且在四季呈現的植栽景觀具有差異，符合第一階段之研究結果：四季植栽景觀變化為影響遊客分布的主要因素；其次的意象為山水風景，常與公路或步道形成複合意象。另外，分析整體國家公園遊客之感知與態度，發現大多遊客在自然環境上感知到正面的意象，而步道設施與社會心理感知中的遊憩活動限制為遊客感知到的主要負面衝擊，因此除了對原有的景觀意象進行保存與保護，需要改善部分之管理策略。整體而言，研究結果可知由VGI鑑別之熱門遊憩據點大致與傳統遊客調查結果相符，且不受調查時間與空間限制；但更能呈現具體空間資訊，作為遊程動線與交通運輸安排及遊憩衝擊管理之依據，同時也包含遊客感知的內涵，可對遊客的體驗進行近一步的規劃與管理。

關鍵字

自發性地理資訊；國家公園；遊憩行為分析；遊客管理

並列摘要

In recent years, with the increasing emphasis on leisure and quality of life, the demand for leisure and recreation has increased day by day. Taiwan has limited resources for recreation with abundant population, which causes the phenomenon of concentrated use of recreation, and national parks have become an important resource of tourism. However, the increasing number of tourists turns into environmental pressure. Therefore, how to balance between ecological conservation and recreational usage, maintaining the quality of recreational experiences and control the environmental impacts caused by tourists at the same time, is a key issue for the long-term management of national park. The solution to the issue is depend on the collection and application of the data. The previous research and survey of tourist behaviors mainly used observations, questionnaires, interviews, etc. The results were greatly limited by time and space and required high labor and economic cost. Recently, the rapidly developed technology of internet and mobile is used generally and changes the way people record and share information. These user-generated contents often contain space-related information, such as Geotag and Check-in tag, called Volunteered Geographic Information (VGI). These data form a bottom-up participation, popularize the information, make information update rapidly, and provide a new direction for the management of national parks. This study attempted to use VGI information uploaded on social media by tourists as data source, helping national park to learn and to manage the tourism behavior. Therefore, there were two main purposes in this study: understanding recreational behaviors, hotspot distribution, and environmental status to assess the impact of recreational behaviors; understanding the imagery of national parks and the attitude perception of tourists. Yangmingshan National Park is located in the Taipei Metropolitan Area. It has convenient transportation and abundant recreational resources that provide urban resident to relax,the number of tourists can reach 19 million people a year. Therefore, this study chose Yangmingshan National Park as a study case with two research stages. Stage 1 was to identify recreational hotspots in the national park. We captured geotagged data of Flickr and conducted Density-Based Clustering Analysis, and conducted on-site survey to know the situation and environment of the hotspots. The travel paths were analyzed by time sequences of data, and we used Association Rules to explore the relationship between hotspots. In stage 2, we conducted content analysis to understand the images of attractions to the public by analyzing the photos uploaded on Instagram. And through the analysis of Facebook comments, we can understand the tourists' feelings and attitudes towards the image. In stage 1, 14 hotspots were identified by VGI, there were Yangming Park, Zhuzihu, Lengshuikeng, Datun Mountain trail, Qingtiangang, etc. These hotspots were roughly corresponding to the past official visitor statistics. According to the time distribution, spring tourists concentrated in Zhuzihu and Yangming Park, while autumn focused on Qingtiangang; the distribution of tourists on the holiday and weekday is more average than past survey; exploring the nighttime data could discover that Datun Mountain as an important nighttime hotspot. The results of on-site survey showed that most of the hotspots had good landscape and had platforms of fences to regulate the behavior of tourists, reducing the occurrence of environment impacts. However, if trails were insufficient in width and had no facilities to regulate the tourists, the behaviors of tourists outside the trail were tend to cause trampling of surrounding vegetation. In addition, the results of association rules showed that the recreational behavior of Yangmingshan National Park were located on the southwest side, including 7 hotspots, and the movements between hotspots were mainly connected by trails, showing the importance of trail management. In stage 2, the main images of national park were animals and plants. The images of vegetation in four seasons were totally different, which is consistent with the results of stage1: the changes of vegetation in four season were the main factors affecting the distribution of tourists. The second image was the landscape, often forming a composite image with roads or trails. In addition, according to the content of tourist comments, it was found that most of the tourists perceived positive image of natural environment, while trail facilities and limits of recreational activities were the main negative impacts that tourists perceived. The results could help preserve or protect the original landscape images, furthermore, indicate the need of improving some management strategies. In conclusion, the results of the study showed that the hotspots identified by VGI were generally consistent with the results of traditional tourist survey, and were not limited by survey time and space. VGI data could present more specific spatial information, which can apply as the basis of travel trail, transportation arrangement, and management of recreational impacts. VGI also contained the contents of tourist perceptions and experiences, which can further apply to the planning and management of recreational experience.