Postdoctoral Research Position (M/F) – Towards Reliable Translation of Natural Language into Formal Specifications for Smart Contracts

id="offer-description" class="ecl-u-type-heading-2 ecl-u-mt-none ecl-u-mb-l">Offer Description The candidate will contribute to the following research activities:. - Design of semantic translation methods: Develop approaches for translating natural language instructions into formal logical formulas. - Development of a Controlled Natural Language: Investigate the design of a language tailored to the application domain, incorporating variables, function calls, and information extracted from the Abstract Syntax Trees (ASTs) of smart contracts. - Syntactic and semantic parsing: Design parsing methods to identify relevant linguistic structures and construct the corresponding semantic representations. - AI-based approaches: Explore and compare different methods, including pretrained language models, reinforcement learning, and latent-variable models. - Integration of symbolic and neural methods: Investigate the complementarity between formal grammatical and semantic approaches and machine learning techniques, with the aim of improving the reliability of translations. - Experimental evaluation: Design and implement evaluation protocols to assess semantic fidelity, logical correctness, robustness, and usability of the proposed methods. Depending on the needs and progress of the project, the research may also explore extending the developed methods to less constrained natural language texts, as well as techniques for ambiguity resolution and interactive clarification of specifications