Author: Ngoc Cat My Tran
Tran, Ngoc Cat My, 2021 Exploring behavioural patterns in interaction data from the “Your Fertility” website, Flinders University, College of Science and Engineering
Terms of Use: This electronic version is (or will be) made publicly available by Flinders University in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. You may use this material for uses permitted under the Copyright Act 1968. If you are the owner of any included third party copyright material and/or you believe that any material has been made available without permission of the copyright owner please contact copyright@flinders.edu.au with the details.
Understanding user’s behavioural patterns while using a certain website can lead to improved website design with more personalized features, or more precisely targeted marketing campaigns. Various papers have been conducted to find the user’s behavioural patterns in the online health platforms, but the majority of them just focus on totally or partially offline behaviours. This study focuses on exploratory data analysis to identify behavioural patterns in using an interactive tool on a fertility education platform (the “Your Fertility” website). A dataset of 4245 people (84% women and 16% men) who voluntarily accessed the online Healthy Conception Tool (HCT) on the “Your Fertility” website was analysed. Depending on the types of variables, the paper used certain statistical techniques including Pearson correlation coefficient, Scatter plot, Chi-Squared Test, Bonferroni correction, One-hot encoding, and Point Biserial Correlation. The paper indicated that the modified profile (the last input of users) has a greater number of correlations than the initial profile (the first input of users), which supports the “linear slider format” issue. Besides, this analysis shows the significant gender differences in behaviours related to age and weight directions, between people having children and no children, and between people having STI (Sexually transmitted infections) test and not. Also, Principal Component Analysis and K-means clustering were applied to figure out the user’s behaviours in groups.
Keywords: user's behaviours, data analysis, Pearson correlation coefficient, Scatter plot, Chi-Squared Test, Bonferroni correction, One-hot encoding, Point Biserial Correlation, Principal Component Analysis, K-means clustering, fertility website, healthcare website
Subject: Computer Science thesis
Thesis type: Masters
Completed: 2021
School: College of Science and Engineering
Supervisor: Dr Shaowen Qin and Dr Richard Leibbrandt