The Department of Tourism, Recreation and Sport Management, University of Florida, Gainesville, Florida, United States of America.
PLoS One. 2018 Nov 2;13(11):e0206820. doi: 10.1371/journal.pone.0206820. eCollection 2018.
This paper uses text data mining to identify long-term developments in tourism academic research from the perspectives of thematic focus, geography, and gender of tourism authorship. Abstracts of papers published in the period of 1970-2017 in high-ranking tourist journals were extracted from the Scopus database and served as data source for the analysis. Fourteen subject areas were identified using the Latent Dirichlet Allocation (LDA) text mining approach. LDA integrated with GIS information allowed to obtain geography distribution and trends of scholarly output, while probabilistic methods of gender identification based on social network data mining were used to track gender dynamics with sufficient confidence. The findings indicate that, while all 14 topics have been prominent from the inception of tourism studies to the present day, the geography of scholarship has notably expanded and the share of female authorship has increased through time and currently almost equals that of male authorship.
本文运用文本数据挖掘方法，从主题焦点、地理和旅游作者的性别这三个视角，来识别旅游学术研究的长期发展趋势。本研究从 Scopus 数据库中提取了 1970 年至 2017 年期间在高排名旅游期刊上发表的论文摘要作为分析的数据源。利用潜在狄利克雷分配（LDA）文本挖掘方法确定了 14 个主题领域。LDA 与 GIS 信息的结合，可以获得学术产出的地理分布和趋势，而基于社会网络数据挖掘的概率性别识别方法，则可以在足够的置信水平上跟踪性别动态。研究结果表明，尽管所有 14 个主题自旅游研究开始至今一直很突出，但学术研究的地理范围显著扩大，女性作者的比例也随着时间的推移而增加，目前几乎与男性作者的比例相当。
