Naslov
Intelligent Public Procurement Monitoring System Powered by Text Mining and Balanced Indicators

Autori
Modrušan, Nikola ; Mršić, Leo ; Rabuzin, Kornelije

Vrsta, podvrsta i kategorija rada
Poglavlja u knjigama, znanstveni

Knjiga
Data Management Technologies and Applications

Urednik/ci
Hammoudi, Slimane ; ‎Quix, ‎Christoph ; Bernardino, Jorge

Izdavač
Springer

Grad
Cham

Godina
2021

Raspon stranica
115-133

ISSN
1865-0929

Ključne riječi
Mining ; Natural Language ; Processing ; Rule Based Decision ; Knowledge Discovery ; Public Procurrement

Sažetak
Optimising the public procurement process through digitalisation has many benefits. In this regard, text mining and emerging technologies used to create intelligent indicators could potentially be applied throughout the public procurement process. Research is showing that these technologies are not yet firmly established in the public procurement process. However, as presented in this paper, they have the potential to transform this process in various ways, and there is a considerable scope as to how this could be achieved. This paper provides insight into the research and corresponding activities of research teams to establish a prototype of a national public procurement monitoring system that can be used as support for policy makers and for educational purposes, to provide best practices, and/or to improve transparency. As part of our research, we created a system that uses text mining techniques to analyse collected data for more than 60.000 cases over 5 years and use that data sample to identify a total of 56 indicator rules. As addition, lessons learned during the research process were used to propose phase-based development system guidelines and infrastructure planning and management. In the final part, several interesting results are presented as showcases that represent potential usability and the ease of use of the system.

Izvorni jezik
Engleski

Znanstvena područja
Informacijske i komunikacijske znanosti

POVEZANOST RADA