University of Twente Student Theses


Discovering behavioral profiles for website visitors of higher educations

Sadeghi, Alireza (2018) Discovering behavioral profiles for website visitors of higher educations.

[img] PDF
Abstract:Several studies made attempts at using behavioural data to support business activities. Some studies attempted to creating user profiles to identify potential customers to increase customer spending and others have attempted to find the best performing algorithm to increase the accuracy of preceding studies. The goal of this paper is to develop a framework for customer profiling and customer attribute prediction within the marketing context using Machine Learning and Customer attributes as well as developing a multi-step user profiling process model for user profiling. The goal in this paper is to discover behavioural profiles of website visitors for higher educations. Previous studies used Machine Learning and customer attributes to identify the most profitable profiles of existing customers for the purpose of increasing spending amount but this paper focuses on identifying behavioural profiles to increase customer base and increase conversion rate. Thus, this study focusses on finding behavioural profiles within website visitors for higher education by utilizing behavioural data and applying proposed model and framework in this paper. The outcome provides insight for University of Twente marketing department as to what behaviours lead to higher conversion. The paper proposes a framework and a model, where the framework provides a guideline for different research goals based on customer attributes & Machine Learning algorithms and the model provides a guideline on the way customer data should be processed to gain a profound insight from data. The analysis reveals three behavioural profiles for the website visitors of the University of Twente by utilizing the framework and the model proposed in this paper. The outcome provides evidence that the outcome is more profound when the proposed framework in combination with proposed model is used compared to the previously one-step user profiling used in the literature.
Item Type:Essay (Master)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:85 business administration, organizational science
Programme:Business Administration MSc (60644)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page