A Case Study of Analyzing User Behavior with Google Analytics and Web Log

• Main Authors: Chang Huang, Chun, 張黃俊
• Other Authors: Cheng, Li-Chen
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/n2j2sv

碩士 === 東吳大學 === 資訊管理學系 === 107 === The purpose of this study was to explore the behavioral patterns of users who successfully contact designers in the upholstery platform. Web Log are combined with GA data to determine the optimal number of clustering through the combination of grouping and dimension reduction. Finally, association rules are used to mine the behavior patterns of users in the within-cluster. The results show that the t-SNE in the dimension reduction method has a more significant effect on the clustering than the PCA. Based on the large itemsets of association rules, the user behaviors in the survey within-cluster are concluded: (1) the number of people is not necessarily proportional to the complexity of the user's behavior, (2) how to let the user quickly find the ideal designer is a key point, (3) brand awareness and website SEO are two major points that must be taken seriously. It is found that the GA tag is complementary to the web page tag, and the GA tag can provide user information that is not recognized in the web page. The results of the study can be further provided to website marketing personnel and managers as an aid to decision making.