Wee Keong Ng

Wee Keong is an expert in machine learning, privacy-preserving techniques, query-permissible encrypted databases, enterprise blockchain systems, data security, and blockchain security. His work is motivated by the need to harness the power of data for the betterment of stakeholders, where the data has confidentiality and privacy issues; where the data may be owned and held by different stakeholders, and where the data is so large that it must be hosted in cloud servers that are never completely secured (susceptible to insider and cyber attacks). He contributes to companies and industries as a technology consultant on projects involving data analytics, artificial intelligence, data privacy and security, and blockchain. He is Associate Editor and Member of Editorial Boards of five journals. Over the years, he has worked with many talented students on their PhD and Master theses.

Current Research and Industry-inspired Projects

• Enterprise Blockchain Systems: Current blockchain systems are not built for enterprise use. Reasons are poor performance, security and vulnerability issues, and the lack of facilities to support enterprise data processing, data analysis and reporting, and data compliance needs. This research looks into the fundamental re-design of blockchain for high transaction throughput, confidentiality and privacy guarantees, the structure of multi-chains, fast consensus algorithms, block compression and archival, blockchain migration, multi-chain connectors, etc. to address enterprise needs for blockchains. We investigate hybrid blockchains - public-private mix of ownership, token and token-less mix of blockchain operations, etc. To support data privacy and advanced data needs, we study the integration of secure multiparty computation with blockchain. This project airtrafficblockchain.io uses our blockchain.
• Blockchain-based Data Protection Platform for Digital Manufacturing Services: We build one of the the world’s first Digital Manufacturing Services (DMS) platform based on Blockchain technology to create digital thread for data protection. As blockchain technology enables transparency in all processes, it will drive the desired behaviour of all parties in the value chain. The data protection framework prevents data theft and data tampering in the 3D printing process. The framework connects designers to Additive Manufacturing Centres (AMC) to simplify the 3D product manufacturing process; secures designer’s data and ensures high quality output. Users are able to place order for 3D printing of parts anytime, anywhere through this framework. This is attractive to large companies managing 10,000s of SKUs and complex global supply chains who want to deploy 3DP, but many not have the infrastructure. We hope to fill the gaps in the 3D printing industry to accelerate industry adoption, making 3D Printing as simple as printing a document from a desktop computer. This company Secur3DP+ uses our blockchain system.
• Privacy Preserving Data Analytics: To realize the full potential of a smart city, one must harness the massive volume of data that have been (and are still being) generated and accumulated in all work and living aspects of the city. A major obstacle in this regard is data privacy; the need to protect Personally Identifiable Information (PII). This research create techniques and systems to address data privacy issues: (1) The need to aggregate data about people from multiple public/private sources for collective data analysis or the application of artificial intelligence techniques without violating privacy; (2) The need to analyze different types of data: numerals, texts, documents, images, etc., which contain sensitive or personal information; (3) The need to understand and manage the data leakage risk of aggregating and analyzing personal data. These companies prismadb.com and newtonis.co use our encrypted database system.
• Deep Learning/Machine Learning Approach to Threat Alert Triage and Risk Modeling There are many indicators of threat in Know-Your-Clients (KYC) processing, from information systems in use to surveillance inputs to open source intelligence to threat typologies. We want to determine the probability of threat (dependent variable) with respect to the linear or non-linear relationships with the many diverse predictor variables (independent variables). This must be done in a real-time continuous manner as KYC is dynamic, including multi-hypothesis tracking. The nature of the relationship among the variables might change with the progression of the time and new types of variables can come into existence that may be proved useful for better prediction accuracy. We use DL/ML approaches to assist financial institutions in triaging KYC threat alerts (with probability and explainable typologies) and to assess risks. With sufficient data, we are can infer new indicators of threats and new typologies.
• Computational Air Mobility: The desire to have air mobility creates a host of computational problems. What is the revenue model for infrastructural air mobility? What is the trade-off between utility provision and safety? What is the relationship between the mobility structure and routing algorithms? For fleet owners, how many aerial vehicles should they own to sustain a profitable operation? What is the pros and cons of having fixed flight schedules versus free flight? We formulate each of these issues as an optimization or constraint satisfaction problem and investigate solution techniques to solve the problems. We also build a multi-agent modeling and simulation system to support empirical investigations.