We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
基于社交媒体数据挖掘的旅游者情绪感知.
- Authors
冯泽琪; 彭霞; 吴亚朝
- Abstract
With the popularization ol smart mobile terminals and social media, a large number of social media data with geo tags have emerged・ The multi-dimensional features of ntexblocation-timen for the social media data make it possible to perceive tourists7 emotions on a fine space-time scale・ In this paper, the data of Sina microblogs released by tourists from 2017 to 2019 are used to analyze the emotions of tourists by BERT model, and the spatial and temporal distribution rules of tourists7 emotions are discussed・ Then based on BERT model, the text classification of tourists7 microblogs is carried out to analyze the emotional characteristics ol tourists under different themes・ Finally, topic extraction is carried out lor tourists7 negative microblogs, and further analysis is made on the related factors that may lead to tourists7 negative emotions. The results showed that the mood of tourists showed diurnal, weekly and seasonal rhythm, and there were differences in the intensity ol emotional response and emotional rhythm for tourists of different genders・ There were differences in the spatial distribution between moderate and strong emotions. Tourists7 microblogs mainly include five themes:"sightseeingn ncateringn "leisure" naccommodationn and nweathern, among which nweather'1 and n catering, T are more likely to generate strong emotions, and tourists7 negative emotions towards weather are often affected by their location and activities. The emotion mining method of tourists proposed in this study can mine the emotion characteristics of tourists from multi-dimensional and multi-level and provide reference for public opinion monitoring and early warning system of tourist destinations.
- Subjects
MICROBLOGS; TOURIST attractions; AFFECTIVE computing; PUBLIC opinion
- Publication
Geography & Geographic Information Science, 2022, Vol 38, Issue 1, p31
- ISSN
1672-0504
- Publication type
Article
- DOI
10.3969/j.issn.1672--0504.2022.01.005