Abstract: In the era of big data,people’s extensive application of social media has provided a new data source for studying tourists’ emotional experiences and exploring their temporal and spatial changes. Incorporating human emotional experiences into geography research is conducive to exploring the laws of human behavior,and investigating the generation of tourists’ emotions and spatiotemporal changes is a vital part of such research. Data combined with geographic locations provide new ideas for studying tourists’ large-scale emotional experiences and their spatiotemporal evolution. Due to the city’s many historical monuments and the plethora of ancient ruins and tombs in its vicinity,tourism has become an important component of the local economy,and the Xi’an region,Shaanxi Province is now one of the most popular tourist destinations in China. This study used Xi’an as the case city,and collected the check-in data of domestic tourists on Weibo over the past three years,using a hot grid map,the Getis-Ord Gi* method,and the X-means clustering method to study the temporal and spatial changes of tourists’ emotional experiences in Xi’an from the dimensions of positive and negative emotion to explore the evolution law of such experiences. The research found as follows：（1） Tourists near the urban center,main axis of the city,main business circles,and scenic spots showed relatively high and stable emotion;the tourist area of Ancient Xi’an,with the Bell Tower and Drum Tower at its center,and the new Qujiang area,which is regarded as the “Reception Room of Xi’an,” were the major areas wherein the tourists displayed high emotional experiences. （2） The transportation hub （i.e.,the station and the entrance and exit of expressways） and marginal areas of the city were the gathering places of negative emotional experiences. （3） Within the past three years,the overall emotion of tourists in Xi’an has not shown much fluctuation. The spatiotemporal clustering of positive and negative emotion showed center-margin patterns,which could be classified into three types：stable,relatively stable,and violently fluctuating. Among these,the stable type was mainly distributed in the city center,near the main business circles,around the traffic trunk lines,and near the scenic spots. The relatively stable area occupied almost all of the Xi’an area,and the violently fluctuating type was mainly observed at the marginal areas of the city. This paper enriches the research methods of tourists’ emotional experiences. Differing from the nuclear density analysis method used in previous studies,this research connects tourists’ emotions and urban areas more accurately and analyzes the law of tourists’ emotional experience distribution and clustering in the city. Moreover,this research holds a certain practical significance for the construction of urban infrastructure,distribution of public service facilities for tourism,and urban and tourism planning. It also plays an enlightening role in including tourists’ emotions in the study of urban planning.

Key words: emotional experience, spatio-temporal, X-means cluster, check-in data