- 旅游
- 文章编号：1009-6000(2019)12-0117-09
- 中图分类号：F592.1 文献标识码：B
- Doi：10.3969/j.issn.1009-6000.2019.12.017
- 项目基金：国家自然科学基金项目青年学科基金项目（71804018）“搜寻匹配视角下中国城市住房市场供需错配机制研究”;教育部人文社会科学研究青年项目（18YJC630039）“不完美信息下的住房租售市场供需匹配机制与管理政策研究”；中央高校基本科研业务费项目（CDJSK190198）“基于项目的价值工程（管理）理论与方法创新研究”；重庆市社科规划办哲学社会科学规划研究项目青年项目（2018QNGL22）“重庆城市‘三生空间’的演变研究”。
- 作者简介：顾渐萍，女，重庆大学建设管理与房地产学院，讲师，博士，主要从事城市与住宅经济学研究；
王远斌，男，重庆大学建设管理与房地产学院，硕士研究生；
刘贵文, 男, 通信作者, 重庆大学建设管理与房地产学院，院长，教授, 博士研究生导师；田宗舜，男，重庆大学建筑管理与房地产学院，讲师，博士，研究方向为数据挖掘。
- 基于文本大数据的游客旅游意象感知挖掘研究 ——以重庆市为例
- Tourism Image Mining Based on User Generated Big Data: Taking Chongqing as An Example
- 浏览量：
- 顾渐萍 王远斌 刘贵文 田宗舜
- GU Jianping WANG Yuanbin LIU Guiwen TIAN Zongshun
- 摘要：
利用旅游者对重庆市景点的旅游评价大数据，通过基于 Word2vec 的词向量模型进行语义挖掘，识别城市旅游意象感知要素类型，结合游客情绪表达分析旅游认知意象和情感意象的对应关系，以及意象感知的空间分异。研究发现重庆市的旅游意象分为空间感知、旅游核心吸引物及旅游支持体系这三个层面以及具体十类意象，游客感知以正面积极为主，但也对交通及旅游服务等因素给予较多负面评价。研究表明在利用 Word2vec 工具处理网络文本基础上对关键词进行聚类，可有效提取意象感知要素，便于分析城市旅游意象结构体系和感知情况，为利用社交媒体大数据实现城市旅游意象感知、指导城市旅游建设提供新的思路。 - 关键词：
自然语言处理；城市旅游意象；语义分析；游客评价；形象差异； - Abstract: Big data from social media plays an important role in sensing and evaluating urban tourism image and enhancing urban tourism competitiveness, user generated contents in social network is a valuable source to extract perceived tourism factors and assess the quality of urban tourism construction. With the increase of data volume, traditional content analysis based on word frequency cannot fully exploit the information provided by these data. Natural language processing based on word vector model was introduced into the study of urban tourism image perception. On the basis of data collecting and preprocessing, word vector model is constructed using the whole corpus, and the emotion of every comment are analyzed. Hot tourism spots are identified according to the location of attractions and the number of comments. Keywords extracted by TF-IDF method are then clustered according to the cosine distance of their word vector. Based on cluster results and tourists’ emotion, the components of Chongqing’s tourism image and elements that caused significant negative emotions were identified. The study finds that Chongqing's tourism image is divided into three aspects: spatial perception, core attractions and tourism support system, and ten specific image factors are recognized: hot tourism spots, urban topography, natural scenery, river and night view, ancient town, recreational facilities, special diet, history and culture, transportation, tourism services. River and night scenes in Chongqing are well-received tourism hotspots. Natural landscape represented by the special geological structure of Wulong area has also left a deep impression on tourists. The evaluation of tourism services is bipolar, and tourists are dissatisfied with transportation, toilets and staff. In addition, the weather also has an important impact on travel experience. This research shows that natural language processing method based on word vector model can effectively draw perceived image elements, which facilitates the use of big data from social media to analyze image perception of urban tourism. This kind of method based on semantics further explores the information provided by user generated contents while reducing the manual element extraction and classification work, which makes the results more scientific and objective.
- Key words: natural language processing; urban tourism image; semantic analysis; user generated content; image difference
期刊浏览
2024 年
2023 年
2022 年
2021 年
2020 年
2019 年
2018 年
2017 年
2016 年
2015 年
2014 年
2013 年
2012 年
2011 年
2010 年
2009 年
2008 年
2007 年
2006 年
2005 年
2004 年
2003 年
2002 年
1999 年