Decentralized Ledger Technology (DLT) based applications employ smart contracts to execute a particular set of rules in a trusted way. Well-designed smart contract languages are Turing-Complete and have deterministic behavior. However, smart contracts, when deployed on the decentralized ledger, are also immutable, thus cannot be changed when faced with a situation unforeseen by developers. Currently, much of the successful utilization of smart contracts are in financial applications where the amounts can be quantified, and the conditions are evident at the outset. But as DLTs are increasingly applied in many non-monetary domains, such as in Internet of Things, or healthcare applications where the variables may not be known in advance, or are dependent on AI algorithms that are not deterministic at the time of smart contract initiation, programming constructs similar to those used in financial applications may be extremely restrictive. Thus, we require better expressivity and flexibility in the contract to account to the fuzziness of such real-world inputs, while ensuring the usual cryptographic guarantees. In this talk, I will describe the work done in the Smart Contracts augmented with Learning and Semantics (SCALES) project that is leading the way to “smarter contracts” that addresses these issues.
When: Friday, September 13, 1-2:00pm
Where: CDM Theater 708
Who: Oshani Seneviratne, Director of Health Data Research, Institute for Data Exploration and Applications (IDEA), Rensselaer Polytechnic Institute (RPI)
Speaker bio: Oshani Seneviratne is the Director of Health Data Research at the Institute for Data Exploration and Applications (IDEA) at the Rensselaer Polytechnic Institute (RPI). At Rensselaer IDEA, Oshani leads the Smart Contracts Augmented with Analytics Learning and Semantics (SCALeS) project. The goal of this project is to predict, detect, and fix initially unforeseen situations in smart contracts that are deployed on the blockchain-based systems utilizing novel combinations of machine learning, program analysis, and semantic technologies. Oshani is also involved in the Health Empowerment by Analytics, Learning, and Semantics (HEALS) Project, where she oversees the research operations targeted at the characterization and analysis of computational medical guidelines for chronic diseases such as diabetes. Prior to Rensselaer, Oshani worked at Oracle specializing in distributed systems, provenance and healthcare-related research. Oshani obtained her Ph.D. in Computer Science from Massachusetts Institute of Technology (MIT) in 2014 under the supervision of Sir Tim Berners-Lee, the inventor ofthe World Wide Web. At MIT, Oshani conducted research on Accountable Systems for the Web, and she is the inventor of a novel web protocol called HyperText Transfer Protocol with Accountability (HTTPA), that has demonstrated its effectiveness in several domains including electronic health care records transfer, and intellectual property protection in web-based decentralized systems.