A carregar...
Publicação
Data-Driven Irregularity Detection in Portuguese Public Procurement
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.87 MB | Adobe PDF |
Orientador(es)
Resumo(s)
Despite being a vital governmental activity that directly affects the quality of life and wellbeing of a country’s citizens, Public Procurement is also highly vulnerable to suffer from
irregularities. Over time, procurement illicit behaviours are becoming more sophisticated
and complex, posing ever so significant challenges for detection and prevention efforts.
However, while that happens, data science and artificial intelligence methods are proving
themselves to be powerful allies in the fight against procurement-related misconduct, due to
the possibilities these methods provide of analysing very large datasets with few costs in less
time and detecting patterns that a human cannot detect on his own. Leveraging these
methods, this research aspires to address these challenges by implementing anomaly
detection techniques tailored to a Portuguese procurement dataset of over two million
records posted to Portal BASE between the years of 2009 and 2022, offering the possibility
of enhancing the efficiency and effectiveness of irregularity detection processes. By
integrating the anomaly score of each contract with red flags that capture hints of suspicious
behaviours from the involved awarding and awarded entities, the study aims to construct a
comprehensive methodology for prioritizing investigations, assigning a score of suspicion or
risk to each contract, which can be used by human auditors to focus their attention and
concentrate their efforts on a much smaller and more manageable set of contracts.
Ultimately, this research seeks to contribute to the advancement of transparency, fair
competition, integrity, and accountability within Public Procurement systems, thereby
safeguarding the interests of citizens and promoting good governance practices.
Descrição
Dissertation presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics, specialization in Data Science
Palavras-chave
Artificial Intelligence Public Procurement Anomaly Detection Transparency Transparency Red Flags SDG 8 - Decent work and economic growth SDG 10 - Reduced inequalities SDG 16 - Peace, justice and strong institutions