應用社群網站於視覺景觀評估之分析

視覺景觀評估被分為專家法與大眾評估法兩大類，於大眾評估法中，遊客自主攝影法則可獲得使用者最直接之感受，以往手機、相機不普及時，發放相機給予受測者之方式，與現今遊客將照片上傳至社群網站上之方式雷同。本研究以Flickr照片分享平台作為研究工具，探討是否上傳照片數量越多之景觀或地點也越受大眾所偏好，以及越偏好或景觀類型越多樣化之地點是否遊客人次也越多。大眾偏好與照片上傳數之變項包括有研究地點、代表性景觀及景觀類型三種，景觀類型多樣性則以Shannon指數計算，遊客人次為研究地點近八年遊客人次之平均。研究方式挑選具有近八年遊客量資料之國家風景區中30處遊憩據點作為研究地點，上傳照片數由Flickr API以地點關鍵字呼叫照片數，刪除各地點以人、食物、寵物為主之照片後，獲得該地點之總上傳照片數；大眾偏好調查方式則挑選每個地點兩個代表性景觀，共60個景觀給予受測者透過網路問卷方式以1至10點填答；景觀類型分類採三角檢測法，將各研究地點景觀分成九大景觀類型，計算出各類偏好平均值與上傳照片數；景觀多樣性指數為各研究地點之景觀類型數量以Shannon指數計算，本研究共收集到384份有效問卷。研究結果為，在Flickr上照片越多之據點與大眾偏好無關，但越偏好之據點遊客人次有越多之趨勢；在代表性景觀方面，大眾越偏好之景觀，上傳照片數量越多，但大眾所偏好之景觀類型不會因為照片上傳數量多寡而變動；遊憩據點景觀多樣性雖然對遊客人次沒有顯著影響，但其分佈大致呈U字型曲線，表示景觀多樣性越高與越低之地點，遊客人次均較多。本研究可提供視覺景觀評估分析一個新的研究工具建議。

關鍵字

Flickr ；遊客自主攝影法；景觀偏好；國家風景區；多樣性指數

並列摘要

In this research, we use photo sharing platform ─ Flickr as the research tool to investigate if the number of photos uploaded at each landscape or location is more preferable by the tourists, and if the landscape is more preferable or having more landscape type causes more tourists to visit. The variables of public preference and number of photos uploaded include research sites, representative landscape, landscape type. The landscape type diversity is calculated by Shannon index; the visit number is the average of the recent eight years at each research sites. In the part of research method, we selected 30 recreation sites within the National Scenic Area which have the data of recent eight years’ visit numbers. The photo uploaded numbers are acquired by Flickr API, calling keywords of each sites and then deleting every photo with their main theme as human, food, pets to obtain total number of photos at each site. The way we do public preference investigation is by online questionnaire survey, letting respondents to rate a total number of 60 representative landscapes (two were selected at each sites) on scale 1 to 10. The triangulation method is adapted in the categorization of landscape types, putting every landscape at each research sites into 9 clusters and calculate means of preference and number of photos uploaded in each cluster. The landscape type diversity index is calculated by Shannon index with the number of landscape types at each research site, recycling effective questionnaire 384. The result shows that the number of photos is not related to public preference, while the number of visits has positive relationship to public’s preference of each site. In the aspect of representative landscape, the more preferable, the more photos people uploaded to Flickr, but public’s preference toward landscape types does not vary with the numbers of photos uploaded. Although the landscape diversity of the recreation sites is not significantly correlated to the number of visits, the distribution appears approximately to be a U-curve, which means whether the landscape diversity is high or low, the number of visits is high. This research can be provided as a new research tool of visual quality assessment.

並列關鍵字

Flickr ； Visitor-employed Photography ； Landscape Preference ； National Scenic Areas ； Shannon Index

參考文獻

1. 大紀元，2014，花蓮年節燈會紅面番鴨亮相，下載日期：2018年5月30日，出處：https://www.epochtimes.com.tw/n81518/%E8%8A%B1%E8%93%AE%E5%B9%B4%E7%AF%80%E7%87%88%E6%9C%83-%E7%B4%85%E9%9D%A2%E7%95%AA%E9%B4%A8%E4%BA%AE%E7%9B%B8.html