E-commerce user activities (dataset 230401_230715)

Published: 24 July 2023| Version 1 | DOI: 10.17632/pvh7jcryx2.1
Adam Wasilewski


The dataset contains information about the collected customer behavior of an e-shop that has implemented a multi-variant user interface serving platform. Data was retrieved from the database with a query: SELECT s.token, s.fingerprint, s.userToken, s.siteId, s.userId, v.url, v.type, v.group, v.category, v.action, v.name, v.time, v.visitKey, v.groupSource FROM `visit` v JOIN session s ON v.sessionId=s.id WHERE v.`time` > '2023-04-01' AND v.`time`<'2023-07-15' The dataset was used to cluster clients and assess the effectiveness of serving different UI variants.


Steps to reproduce

In the first step, data from 73 days were used to cluster the customers. The analysis was performed using the agglomerative clustering method, with 4 resultant clusters set. The second part examined e-commerce performance indicators (CR, AVB and PCR) for each cluster of users. The clusters were additionally divided internally (randomly, 50% each) to determine the study group and the comparison group.


Intelligent User Interface, Behavioral e-Commerce, Cluster Analysis


Narodowe Centrum Badań i Rozwoju