Visual Analysis Of Destination Image Comparison Based On Tourism UGC

A wide variety of tourism platforms have accumulated a large amount of UGC(User Generated Content)data,which contains users' records,comments and experiences during the travel period.Tourists' perceptions and viewpoints about the destination in UGC are called destination images in the field of tourism.Tourism managers need to find out the differences among destinations by studying comparative of destination image and exploring the evolution trend of destination image over time,to further provide desion supports for themselves to improve the competitiveness of destinations.At present,destination image studies based on tourism UGC mainly focus on the analysis or construction of single destination image.There are few image comparative analysis studies,and the impact of time dimension on destination image has not been fully considered.In light of the above problems,based on tourism UGC data,this thesis conducts research on the image comparison of destinations by using visual analysis methods,to meet the requirements of tourism managers.In summary,the contributions of this thesis are as follows:1)Construct a visual analysis model of destination image comparison based on tourism UGC.Firstly,the model extracts destinations,keywords,emotions and temporal information from tourism UGC data.Then,visualization methods and technologies such as time sequence visualization,sentiment visualization,keyword visualization,visual encoding and interactive control are applied in the model.The images of multiple destinations are compared from two levels of overview and detail,so as to find the image differences in the five dimensions,such as atmosphere,service and food.2)Propose a multi-category sentiment visualization method for the overview and time series comparison of destination image collections.In order to achieve the overview level and time-series comparative analysis of the multiple destinations image,and guide users to select the destination of interest for detailed analysis,this method combines the dimensional image unit and the overall image unit.At the same time,this method uses interactive methods,such as reconfiguration,prompting and selection,to realize intuitive analysis of the cognition and sentiment of the destination image.Further,this methd combines the time dimension of UGC data to realize the time series evolution analysis of the image.3)Propose a radar chart based on keyword visualization method for detailed comparison of destination image collections.In order to achieve the detailed comparison level of the five dimensions of destination images,this method integrates and refines radar chart visualization and text keyword cloud visualization technology.At the same time,this method uses interactive methods,such as layer addition,selection and reconfiguration,to achieve the exploration and analysis of image from the five cognitive dimensions.Furthermore,the effects of different cognitive dimensions on the destination image are found by this method,and the multi-attribute comparison of the destination images are realized in detail.Based on the above contributions,the prototype system of "visual analysis of destination image comparison based on tourism UGC" is designed and implemented.Through multi-view linkage and rich flexible interactions,the system can help tourism managers to determine the destination comparison candidate set and complete the image comparison of tourism destinations from two levels of overview and detail.Tourism managers can discover and analyze the differences from various dimensions,such as atmosphere,service and food,and explore the temporal evolution trend of the image.Finally,the usability and effectiveness of the methods and the system proposed in this thesis are verified by case study and expert evaluation.The insights of this thesis can help tourism managers to analyze the comparison of destination images better and enhance the competitiveness of the destinations,so as to provide image support for managers to enhance the competitiveness of the destination.