1. To participate please register first

2. Accept Terms & Conditions on first use

3. Staff must select Single Sign On to login

Idea Details

Stage: Phase 2

Submission Team
Vote Score:1
Unique Views:10
Total Views:2,744
Linked Submissions:
No Linked Submissions
This Proposal is hidden. Change Visibility
BIG DATA Analytics in Auditing to Identify Corruption Risks in Public Procurement
Vote Score

Supreme Audit Institutions (SAIs) are pivotal players in the fight against corruption in the public sector. Audits make risks visible, identify misuse of public finances, abuse of power for private gain that harms public services delivered, thus diminishes taxpayers' well-being. SAIs, via reporting their audit findings and accompanying recommendations, contribute to a climate of transparency that has positive impact on building robust and effective systems contributing to preventing and tackling corruption leading to good governance.

State Audit Office of Georgia (SAOG) representing a SAI of Georgia strives to further advance its actions against corruption through sophisticated examination of public procurement. SAOG can take an advantage of big data analysis and machine learning to further advance identification of corruption risks in public procurement and examine these risks by audit procedures.

Particularly, SAOG aims to join and analyze big data from various sources that are currently scattered across different organizations and platforms, such as: State Procurement Agency's electronic platform, State Treasury, Public Registry, Register Base of Companies, Electronic Catalog of Purchasing Facilities and Suppliers and etc.

This data would be structured and used in conjunction with the existing data, information and knowledge owned by SAOG, accumulated throughout the years during the audits of state entities, to build algorithm, identifying corruption in procurement. Namely, based on the revealed deficiencies in public procurement, there would be elaborated Machine Learning model that would enable to uncover hidden patterns, unknown correlations to identify risky areas in the whole population of the procurement database. As a result, it will provide opportunity to identify specific deviations, in more resource efficient manner, making it possible to increase the audit coverage of public procurement much higher and enhance SAOG's contribution to good governance.

Data collection and processing should be automatized and model underpinned would assess the risks on a regular basis, as they might change over time with corrupt actors trying to find new loopholes as old schemes are being disclosed through successful audit processes and corruption risks mitigation strategies as a response to audit recommendations. More the model is trained and more audit findings are supplied, better the algorithm would predict risky procurements, hence, there would be established a process of continuous improvement of the model logically resulting in more advanced and precise identification of the patterns related to public procurement.

In addition to that, risks and systemic deficiencies identified in procurement will be illustrated on the analytical platform - Budget Monitor (www.budgetmonitor.ge) in dynamic and user-friendly visualizations, such as network diagrams and interactive dashboards. Budget Monitor is a thrice international award winning, citizen engagement ICT tool, developed by SAOG - to increase citizen oversight and enhance public control on the usage of taxpayers' monies. On the one hand, it provides comprehensive budgetary information to citizens, using easy to understand data visualizations. On the other hand, it enables citizens to participate in audit process - submit information about corruption risks in public finances and support its elimination. Budget Monitor was awarded Public Participation in Fiscal Policy and Budget Making Award by Global Initiative for Fiscal Transparency (GIFT); United Nations World Summit Award of digital innovations in the category of the Government & Citizen Engagement; Georgia's IT Innovation (GITI) award in the category of the Best Online Information Resource.

Expertise Required
Audit Big Data Analysis Data Visualization Machine Learning Public procurement analys 



No evaluations ready
Recorded Outcomes
Record an Outcome