Along the trend of digital transformation, brand website does not only promote brand image, but it also provides an experience in the path-to-purchase. With the incorporation of ecommerce capabilities on brand websites, we will need to revisit how we do website analytics to suit the new needs. The increased online usage allows for big data analysis with real-time behavioural data. This paper presents a new methodology for web analytics. Based on a case study of real-life brand store website in Hong Kong, the exploration study employs the new use of machine learning technique – UX2Vec, an unsupervised learning methodology for vectorization and embedding, plus k-Means clustering algorithm to discover insights and suggest improvement. An analysis was performed with 30-day data of 69,648 page view records gathered from the website. The proposed method led to a successful clustering result to characterize the website based on machine learning technique.
|Title of host publication||WI-IAT '21: IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology|
|Editors||Xiaoying Gao, Guangyan Huang, Jie Cao, Jian Cao, Ke Deng|
|Publisher||Association for Computing Machinery (ACM)|
|Number of pages||5|
|Publication status||Published - Dec 2021|
Scopus Subject Areas
- Social Sciences(all)