img

官方微信

高级检索

中国沙漠 ›› 2026, Vol. 46 ›› Issue (4): 12-21.DOI: 10.7522/j.issn.1000-694X.2025.00160

• • 上一篇    下一篇

基于用户生成内容(UGC)数据的柴达木盆地雅丹地貌旅游形象感知分析

郜学敏1(), 董治宝2, 李豪3, 刘志芳4   

  1. 1.晋中学院 旅游管理系,山西 晋中 030619
    2.陕西师范大学 地理科学与旅游学院,陕西 西安 710119
    3.太原师范学院,经济与管理学院,山西 晋中 030619
    4.太原师范学院,地理科学学院,山西 晋中 030619
  • 收稿日期:2025-09-14 修回日期:2025-10-27 出版日期:2026-07-20 发布日期:2026-08-27
  • 作者简介:郜学敏(1987—),女,山西晋城人,博士,副教授,主要从事旅游地理、旅游规划开发等的研究。E-mail: waipoqiaoqt@163.com
  • 基金资助:
    山西省哲学社会科学规划课题(2023YY296)

Research on tourism image perception of yardang landform in Qaidam Basin based on UGC data analysis

Xuemin Gao1(), Zhibao Dong2, Hao Li3, Zhifang Liu4   

  1. 1.Department of Tourism Management,Jinzhong University,Jinzhong 030619,Shanxi,China
    2.School of Geography and Tourism,Shaanxi Normal University,Xi'an 710119,China
    3.School of Economics and Management /, Taiyuan Normal University,Jinzhong 030619,Shanxi,China
    4.School of Geographical Sciences, Taiyuan Normal University,Jinzhong 030619,Shanxi,China
  • Received:2025-09-14 Revised:2025-10-27 Online:2026-07-20 Published:2026-08-27

摘要:

柴达木盆地雅丹地貌旅游的发展,长期受相对偏远的地理区位制约,且与其他地方的旅游形象缺乏区分度,市场辨识度较低,因此,科学重塑其旅游形象已迫在眉睫。本文运用Python数据挖掘技术采集国内主要旅游平台2019年1月至2025年7月的用户生成内容(UGC)文本数据,利用TF-IDF算法、LDA主题模型等自然语言处理技术(NLP)从认知-情感-整体3个维度分析柴达木盆地雅丹地貌旅游感知形象。结果表明:(1)最高TF-IDF值集中于地貌特征名词“雅丹”(0.052 438)“水上”(0.032 429),认知形象包括地貌旅游资源、旅游评价、旅游体验、旅游设施服务和旅游活动5个主题维度,“水上雅丹”“类火星地貌”是核心认知形象要素;(2)柴达木盆地雅丹地貌旅游情感极性中,积极情感占比最高(79.2%),水上雅丹地貌景观是其积极情感的主要来源,消极情感占比较低(15.3%),集中在旅游设施不足、服务缺位和管理混乱等方面,中性情感占比最少(5.5%);(3)整体形象呈核心层-中间层-边缘层3层结构,核心层形象是“水上雅丹”,中间层形象是“西北自驾圣地”,边缘层则包括“最像火星的地方”“魔鬼城”等延伸形象。

关键词: 雅丹地貌, 旅游形象, 用户生成内容(UGC), 柴达木盆地

Abstract:

The tourism development of the yardang landform in the Qaidam Basin has long been constrained by its relatively remote geographical location. Simultaneously, its tourism image lacks distinctiveness compared to other yardang destinations, resulting in low market recognition. Therefore, it has become imperative to scientifically reshape its tourism image. This study employs Python-based data mining techniques to collect user-generated content (UGC) from major domestic travel platforms (January 2019 to July 2025). Using natural language processing (NLP) technologies such as the TF-IDF algorithm and LDA topic modeling, the tourism perception image of the Qaidam Basin yardang landform is analyzed across three dimensions: cognitive image, affective image, and holistic image. The results indicate: (1) The highest TF-IDF values are concentrated on geomorphological feature nouns, namely “yardang” (0.052 438) and “water” (0.032 429). The cognitive image comprises five thematic dimensions: geomorphological tourism resources, tourism evaluation, tourism experience, tourism facilities and services, and tourism activities, with “water yardang” and “Mars-like landscape” being the core cognitive elements. (2) In the sentiment analysis of tourism perceptions for the yardang landform in the Qaidam Basin, positive emotions accounted for the highest proportion (79.2%). The water yardang landscape serves as the primary source of these positive sentiments. Negative emotions constituted a relatively small proportion (15.3%) and were mainly concentrated in areas such as inadequate tourism facilities, insufficient services, and disorganized management. Neutral emotions represented the smallest share (5.5%). (3) The holistic image exhibits a three-tier structure: the core layer is “water yardang”, the intermediate layer is “the holy land for road trips in Northwest China”, and the peripheral layer includes extended images such as “the most Mars-analog place on Earth” and “Devil's City”.

Key words: yardang landform, tourism image, user-generated content (UGC), Qaidam Basin

中图分类号: