基于“两步路”平台的南岭徒步旅行者旅游凝视近景偏好分析

An Analysis of Tourist Gazing Preference of Nanling Hikers’Based on 2bulu.com

以南岭山地湖南省区域(越城岭、都庞岭、萌渚岭、骑田岭)为研究地,基于Python制作网络爬虫,提取该区域徒步旅行者在“两步路”平台发布的图片标记及其基本信息,共获取2016—2020年有效标记照片7857张。采用计算机深度学习算法解析图片表征内容,并将解析结果与其对应信息导入Nvivo 12.0质性分析软件进行机器编码,对图片元素分类整合,采用图片内容分析法并结合扎根理论,对标注照片所呈现的景观进行再次分类,提取高频词及利用节点聚合等功能,分析徒步旅行者对湖南省南岭山地的近景标注偏好。研究表明:南岭山地湖南区域都庞岭、越城岭、萌渚岭、骑田岭的户外旅行者对近景标注元素总体上以山岳景观为主,对峡谷、溪流等自然要素的关注度较高,而对作为森林主体之植物的关注度不高,可能与徒步旅行者的生物基础知识水平有关;不同区域对近景标注图片反映的内容呈现出一定的差异性,这些差异缘于主要徒步路线经过的自然地理环境、辨识度高的地理标志以及当地的开发与管理措施等不同;同一区域的近景关注对象随着时序发展而逐渐丰富,曾出现过的近景标注元素会逐渐覆盖该区域所有元素类型,因此新出现的近景标注元素的数量会越来越少,这说明旅行者近景标注的偏好相似度会越来越高;不同年份之间,因管理措施、开发建设、国家政策的不同,旅行者对近景标注的偏好会有所差异。

Taking the Nanling Mountainous Region of Hunan Province(Yuechengling,Dupangling,Mengzhuling,Qitianling)as the research site,a web crawler is made based on Python,the pictures posted by hikers in this area on 2bulu.com are extracted,and a total of 7857 valid tagged photos from 2016 to 2020 were obtained.The computer deep learning algorithm is used to analyze the image representation content,and the analysis result and its corresponding information are imported into the Nvivo 12.0 qualitative analysis software for machine coding,the image elements are classified and integrated,and the image content analysis method combined with grounded theory is used to present the annotated photos.The landscape is classified again,highfrequency words are extracted,and node aggregation is used to analyze hikers’preference for close-range labeling of Nanling Mountains in Hunan Province.The results show that:Outdoor travelers in Dupangling,Yuechengling,Mengzhuling and Qitianling in the southern Hunan region of Nanling Mountains generally pay more attention to the near-view labeling elements as mountain landscapes,and natural elements such as canyons and streams are more concerned.However,the low level of attention to plants as the main part of the forest may be related to the level of basic biological knowledge of outdoor hikers;different areas show certain differences in the content reflected by the close-range annotation pictures,and these differences are due to the main hiking routes passing by.The natural geographical environment,highly recognizable geographical indications,and local development and management measures are different;the close-range attention objects in the same area gradually become richer with the development of time sequence,and the close-range label elements that have appeared will gradually cover all element types in the area.Therefore,the number of newly emerging close-range annotation elements will become less and less,which means that the preference similarity of travelers'close-range annotations will become higher and higher;between different years,due to differences in management measures,development and construction,and national policies,travelers.The preference for close-range labeling will vary.