An exploratory analysis of blockchain technology research in humanitarian supply chains and logistics using the Latent Dirichlet Allocation based topic modelling approach

Type of content
Theses / Dissertations
Publisher's DOI/URI
Thesis discipline
Economics
Degree name
Master of Philosophy
Publisher
Journal Title
Journal ISSN
Volume Title
Language
Date
2023
Authors
Patil, Shriniket
Abstract

In humanitarian relief operations, the logistics aspect accounts to approximately 80% of the effort (Ko & Verity, 2018; Trunick, 2005; Van Wassenhove, 2006) wherein the sector experiences significant challenges related to the last mile distribution, transparency, collaboration, accountability, trust, information sharing, time, cost and resilience (Aranda et al., 2019; Cozzolino, 2012; Dubey et al., 2020; Kovács & Spens, 2007; L'Hermitte & Nair, 2020; Negi & Negi, 2020). Extant literature suggests that the academia and industry have extensively researched the use cases and applications of emerging blockchain technology in commercial supply chains and logistics, and identified promising benefits such as trust, transparency, traceability, immutability, provenance, disintermediation and compliance for enabling its adoption (Kshetri, 2018; Min, 2019; Nandi et al., 2021; Niranjanamurthy et al., 2019; Sharma et al., 2019; Wang et al., 2019). However, corresponding blockchain technology led research initiatives in the context of non-commercial humanitarian supply chains have remained extremely scant.

Subject matter experts, noted academics and specialists in the humanitarian sector have been calling for emerging technology led and interdisciplinary research initiatives in this scantly explored area (Aranda et al., 2019; Baharmand & Comes, 2019; Coppi & Fast, 2019; Dubey et al., 2020; Keenaghan et al., 2019; Zwitter & Boisse-Despiaux, 2018). Taking into consideration the research gap as well as the strong endorsement from the expert humanitarian logistics community, the study at hand explores the blockchain technology research literature in humanitarian supply chains and logistics operations by employing a generative probabilistic Latent Dirichlet Allocation (LDA) topic modelling technique.

A predetermined dataset comprising 54 full-text documents pertaining to blockchain technology research in humanitarian supply chains and logistics is analysed using the LDA topic modelling method to reveal three i.e., k = 3 topics – Technology, Organisational Operations, Systems Adoption. The LDA topic model also uncovers pertinent key words such as trust, transparency, coordination, traceability, smart contracts, cost, time, communication, coordination, and food which are consistent with the factors confirmed in the previous and ongoing research initiatives on blockchain technology applications in humanitarian supply chain and logistics.

Next, a comparative analysis between the manual thematic analysis and the machine learning based topic model illustrates that in case of a successful LDA topic extraction process, there is ample scope to integrate the two methods and utilise them as a concurrent mixed methods strategy to inform subsequent qualitative and quantitative research initiatives. Correspondingly, based on the LDA topics and thematic analysis of the literature, researchers may contemplate on diversifying and extending the extant socio-technical information systems theories such as the TOE Framework (Baker, 2012; Tornatzky et al., 1990) with state-of-the- art macro and micro level theories such as the Mikropolis model (Wahoff et al., 2012) for advancing the knowledge in the context of blockchain technology and humanitarian supply chains and logistics studies. And lastly, the application of the LDA topic model in conjunction with the thematic analysis approach and literature reviews may demonstrate the viability of a unique type of a secondary mixed methods approach.

Description
Citation
Keywords
Ngā upoko tukutuku/Māori subject headings
ANZSRC fields of research
Rights
All Rights Reserved