Investigate organizational member engagement through financial x-ray and artificial neural networks

Hason Rudd, David

Investigate organizational member engagement through financial x-ray and artificial neural networks

Hason Rudd, David

Permalink

Publication Type:: Thesis
Issue Date:: 2023

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download thesisAdobe PDF (15.02 MB)

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Hason Rudd, David
dc.date.accessioned	2025-01-24T01:54:23Z
dc.date.available	2025-01-24T01:54:23Z
dc.date.issued	2023
dc.identifier.uri	http://hdl.handle.net/10453/184155
dc.description	University of Technology Sydney. Faculty of Engineering and Information Technology.	en_US.UTF-8
dc.description.abstract	Comprehensive understanding of member engagement and churn is imperative for employees within financial institutions and associations, necessitating shifting from conventional approaches toward more sophisticated analytical paradigms. Employing customer voice (CV), financial literacy (FL), and customer relationship management (CRM) data for churn analysis helps to define member engagement level, facilitating effective retention strategies and fostering long-term loyalty for sustained growth. Members with improved financial knowledge are better equipped to make advantageous decisions and less likely to churn due to misconceptions or unmet expectations. Concurrently, in an era where Telephonic interactions have become the norm post-COVID-19, and emotional content derived from such conversational interactions currently provides real-time insights into member’s sentiments, and can be utilized as a predictor for churn modeling. Although many previous studies have explored helpful information to analyze member’s behavior for churn, they often overlooked bridging member engagement and churn through a holistic view of members’ interactions, emotions, FL, and CRM data. Several approaches to addressing these issues have been introduced using single data sources, e.g., transactional, demographic, and textual data, which are not multifaceted views of member behavior. Current efforts are limited to three main challenges. First, transactional data employed in several recent studies only reflected prediction outcomes, rather than experience or underlying causes for churn. Second, although demographic data have been employed in many studies; static data does not capture dynamic customer satisfaction. Third, social media data (textual data) lacks personalized insights, such as voice interaction and financial skills. Therefore, this thesis leverages a multimodal modeling approach to capture multifaceted insights for member engagement. The main themes of this thesis include 1. Introduce novel speech emotion recognition (SER) methods, developing a VGGoptiVMD algorithm to capture real-time emotions from CV data, enabling early detection of dissatisfaction and personalized interactions. 2. Develop an SER model using CV signal processing, harmonic and percussive components from the Mel Spectrogram. 3. Develop SMOGN-COREG semi-supervised machine-learning techniques to extract patterns from unlabeled financial network data. 4. Develop a causal model to understand root causes for churn in member-centric organizations. 5. Develop a multimodal hybrid fusion learning model integrates FL metrics, behavioral indicators, and voice emotional features. This study introduces a multimodal hybrid fusion model that combines CV, FL, and CRM data to enhance understanding of member engagement for churn risk analysis. It offers a comprehensive framework to improve member engagement and reduce churn, providing a strategic blueprint for sustainable growth in CRM research.	en_US.UTF-8
dc.format	Thesis (PhD)
dc.language.iso	en_US	en_US.UTF-8
dc.relation	https://opus.lib.uts.edu.au/bitstream/10453/184155/1/thesis.pdf
dc.rights	info:eu-repo/semantics/openAccess
dc.rights	The author owns the copyright in this thesis including all reproduction and reuse rights for the work. The work may not be altered without the permission of the copyright owner. Attribution is essential when quoting or paraphrasing from this thesis.
dc.rights	© 2023 David Hason Rudd
dc.rights	au.edu.uts.lib/cph
dc.title	Investigate organizational member engagement through financial x-ray and artificial neural networks	en_US.UTF-8
dc.type	Thesis
utslib.copyright.status	open_access	*

Abstract:

Comprehensive understanding of member engagement and churn is imperative for employees within financial institutions and associations, necessitating shifting from conventional approaches toward more sophisticated analytical paradigms. Employing customer voice (CV), financial literacy (FL), and customer relationship management (CRM) data for churn analysis helps to define member engagement level, facilitating effective retention strategies and fostering long-term loyalty for sustained growth. Members with improved financial knowledge are better equipped to make advantageous decisions and less likely to churn due to misconceptions or unmet expectations. Concurrently, in an era where Telephonic interactions have become the norm post-COVID-19, and emotional content derived from such conversational interactions currently provides real-time insights into member’s sentiments, and can be utilized as a predictor for churn modeling. Although many previous studies have explored helpful information to analyze member’s behavior for churn, they often overlooked bridging member engagement and churn through a holistic view of members’ interactions, emotions, FL, and CRM data. Several approaches to addressing these issues have been introduced using single data sources, e.g., transactional, demographic, and textual data, which are not multifaceted views of member behavior. Current efforts are limited to three main challenges. First, transactional data employed in several recent studies only reflected prediction outcomes, rather than experience or underlying causes for churn. Second, although demographic data have been employed in many studies; static data does not capture dynamic customer satisfaction. Third, social media data (textual data) lacks personalized insights, such as voice interaction and financial skills. Therefore, this thesis leverages a multimodal modeling approach to capture multifaceted insights for member engagement. The main themes of this thesis include 1. Introduce novel speech emotion recognition (SER) methods, developing a VGGoptiVMD algorithm to capture real-time emotions from CV data, enabling early detection of dissatisfaction and personalized interactions. 2. Develop an SER model using CV signal processing, harmonic and percussive components from the Mel Spectrogram. 3. Develop SMOGN-COREG semi-supervised machine-learning techniques to extract patterns from unlabeled financial network data. 4. Develop a causal model to understand root causes for churn in member-centric organizations. 5. Develop a multimodal hybrid fusion learning model integrates FL metrics, behavioral indicators, and voice emotional features. This study introduces a multimodal hybrid fusion model that combines CV, FL, and CRM data to enhance understanding of member engagement for churn risk analysis. It offers a comprehensive framework to improve member engagement and reduce churn, providing a strategic blueprint for sustainable growth in CRM research.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/184155