Service Needs

Reference No.

N-0043

Title

Automate the verification of electronic documents

Business Needs / Challenges

Around 70,000 to 80,000 electronic documents were received through the system every day and about 2,000-3,000 of them were suspected to have inaccurate information included or selected for manual check.  For those suspected with inaccurate information document, officers concerned need to follow up by contacting the submitter through telephone or query through the system for clarification and correction.  For those documents suspected to involve deceptive information, the case would be referred to law enforcement agencies for investigation.  The project aims to enhance the existing system using artificial intelligence, machine learning and other technologies to identify and process the incoming electronic documents with suspected inaccurate information.

Application Areas

Commerce and Industry

Expected Outcomes

With the application of artificial intelligence, machine learning, robotic process automation and other technologies, the enhanced system should be able to minimise human effort for manual checking by improving the accuracy of the received documents (such as through automatic confirmation and clarification of dubious information with the submitters during submission), identifying documents with suspected inaccurate information in a more targeted and precise manner, and improving the efficiency in the manual checking process (such as through use of automatic or machine-assisted checking and correction).

Technologies to be Used

Artificial Intelligence (AI)
Robotic Process Automation