Research Grant Call

• Advance the science of trust technologies, and
• Partner industry to explore and solve systemic real-world problems that once solved, can be scaled for other players in the sector or industry. 

---

Scope of Research

• Searchable Symmetric Encryption (SSE): i) Leak user access / search patterns to servers, and ii) Insufficient complex query expressiveness for search on encrypted data.
• Full Homomorphic Encryption (FHE): Performance issues due to algorithm complexity and large ciphertext sizes which lead to high computational intensiveness and communication overhead and latency. 

• Private Set Intersection (PSI): i) Performance and scalability issues when the number of parties involved increases, and ii) limited expressive queries and functionalities (e.g. lack of support for range, threshold, and conjunctive queries).
• Multi-Party Computation (MPC): Practicability and scalability issues as protocols are communication and computationally intensive, particularly when dealing with a large number of parties or complex computations.
• Differential Privacy (DP): How to strike a balance between privacy and utility when applying DP.
• Federated Learning (FL): How to develop mechanisms for collaborative model selection, model aggregation, and model update strategies in FL minimizing impact on training accuracy.
• Data Synthesis (DS): How to handle complex data types, capture spatio-temporal dependencies, and incorporate privacy-preserving mechanism in data synthesis process.
  

Trusted Digital ID and Web 3.0

• Trusted Digital Identity Network with emphasis on credentials and verification for digital ID, digital wallet and credential management, and digital ID network reference architecture.
• Tokenisation and token standards for digital assets to support security and auditing.
• Biometric-assisted authentication in Web 3.0.
• Trust in Web 3.0 applications through blockchain traceability and analytics, risk assessments and verification of smart contracts against legal contracts and the use of self-sovereign identity (SSI).
  

Trusted Compute

• Design and analysis of system environment for trusted compute based on applications of TEE.
• Verifiable Computing (VC) for the correctness in distributed, decentralised infrastructure for specific computation tasks, e.g. e-voting tally.
• Digital Evidence: An important part of digital trust is the ability to establish accountability and liability in case of fraud, transaction failure or abnormal behaviour. To achieve this, digital evidence borrows digital forensic concepts and techniques that are relevant in a holistic approach to digital trust.
  

Trusted Accreditation

• Hallucination – Tackle factually inaccurate or fake outputs in Generative AI.
• Fairness – No unintended bias: AI system makes same decision even if an attribute changes, and data used to train model is representative.
• Explainability – Explain behaviour of AI models and/or multi-modal models to understand how the inner mechanics. impact the generated output. Examples of problem include lack of transparency, non-deterministic outputs, high dimensionality, and overfitting.
• Robustness – Address issues relevant to robustness e.g., include non-adversarial robustness and privacy related attacks.

Full details of the scope for the grant call can be found in the DTC Research Grant Call Rules and Guidelines.

Evaluation Criteria

Proposals should clearly state the following:

• Alignment of proposal to DTC's objectives and direction. 
• Explaining novelty of research and the needle-moving research challenge that the proposal will solve.
• Explaining potential industry application or impact. 
• Justifying benefit of the research to digital trust.
• Explaining relevance of research to Singapore.

---

Contact Information

For any queries, please contact [email protected].