Abstract
Purpose: The purpose of this study is to develop a country logistics performance assessment methodology and tool based upon a textual data analytics approach.
Design/methodology/approach: The study uses a design science approach to develop and train the prototype performance classification tool from a text corpus of country logistics assessments. Data were collected from 21 text documents from 20 to 60 pages, describing the logistics systems of 20 countries from 2006 to 2014. The extracted texts were tokenised with Natural Language Toolkit (NLTK) and Python programming language, and techniques such as Keyword Analysis, Word Frequency Analysis, Collocation Analysis, and Text Classification are employed.
Findings: The methodology is able to generate results in the form of logistics performance assessments that can complement the existing performance assessment systems such as the Logistics Performance Index (LPI). The analysed text corpus, CSCMP Global Perspectives, is found to contain largely positive and descriptive information about the analysed countries. It is found that a deep involvement of both the logistics domain-expert and the text analyst is required in order to produce usable knowledge and a robust methodology.
Practical implications: When successfully implemented the developed methodology can be used by managers and government bodies such as the World Bank and its stakeholders to complement the existing assessment methods such as the Logistics Performance Index (LPI).
Originality/value: The methodology developed in this paper is new and unconventional for logistics system assessment and appraisal. The research can help enhance the operational and strategic business intelligence capabilities of organisations by creating a new, additional source of knowledge on country logistics performance for decision makers. The paper also demonstrates the potential of textual data analytics in supply chain management as an alternative big data application.
Design/methodology/approach: The study uses a design science approach to develop and train the prototype performance classification tool from a text corpus of country logistics assessments. Data were collected from 21 text documents from 20 to 60 pages, describing the logistics systems of 20 countries from 2006 to 2014. The extracted texts were tokenised with Natural Language Toolkit (NLTK) and Python programming language, and techniques such as Keyword Analysis, Word Frequency Analysis, Collocation Analysis, and Text Classification are employed.
Findings: The methodology is able to generate results in the form of logistics performance assessments that can complement the existing performance assessment systems such as the Logistics Performance Index (LPI). The analysed text corpus, CSCMP Global Perspectives, is found to contain largely positive and descriptive information about the analysed countries. It is found that a deep involvement of both the logistics domain-expert and the text analyst is required in order to produce usable knowledge and a robust methodology.
Practical implications: When successfully implemented the developed methodology can be used by managers and government bodies such as the World Bank and its stakeholders to complement the existing assessment methods such as the Logistics Performance Index (LPI).
Originality/value: The methodology developed in this paper is new and unconventional for logistics system assessment and appraisal. The research can help enhance the operational and strategic business intelligence capabilities of organisations by creating a new, additional source of knowledge on country logistics performance for decision makers. The paper also demonstrates the potential of textual data analytics in supply chain management as an alternative big data application.
| Originalsprog | Engelsk |
|---|---|
| Publikationsdato | 2019 |
| Antal sider | 16 |
| Status | Udgivet - 2019 |
| Begivenhed | The 31st NOFOMA Conference 2019: Annual Nordic Logistics Research Network Conference: Supply Chains and Sustainable Development of Societies - BI - campus Oslo, Oslo, Norge Varighed: 12 jun. 2019 → 14 jun. 2019 Konferencens nummer: 31 https://www.bi.edu/about-bi/calendar/2019/june/nofoma2019/ |
Konference
| Konference | The 31st NOFOMA Conference 2019 |
|---|---|
| Nummer | 31 |
| Lokation | BI - campus Oslo |
| Land/Område | Norge |
| By | Oslo |
| Periode | 12/06/2019 → 14/06/2019 |
| Internetadresse |
Emneord
- Logistics
- Performance Assessment
- Trade Facilitation
- Logistics Performance Index (LPI)
- Computer Aided Text Analysis (CATA)
- Machine learning
- Big Data