Principal Research Scientist, Symbolic AI (1-year contract), SOL
You may be asked to participate in the development of use cases and delivery of proof-of-concept prototypes for government and industry partners in the areas of computable contracts and "rules as code". Experience in requirements analysis and agile, iterative methodologies will be useful at this stage. The work of the team will be largely determined by these use cases and customers. We will work with multiple use cases so that we can abstract across them and develop generalized, reusable infrastructure. We will work with specific customers so we can build useful applications on top of that infrastructure. You will contribute to the design and development of open-source software written to serve those use cases. You will work as part of a team of research engineers, developers, and domain expertise and you will be expected to actively participate in team meetings and text-based communications. You will be expected to be active members of the applied research community, by validating and contributing to advances in computational law.
You may be asked to conduct use-inspired research at the direction of the Principal Investigator. This includes
- reading academic papers, standards publications, and related material such as blog posts and presentations;
- coming up to speed in fields which may be new to you, such as computational linguistics, mathematical logic, programming language design, formal verification, and model-based architecture;
- reading (and updating/commenting, relevant) existing software documentation;
- creating presentations, video recordings, walkthroughs, etc. to document and explain research work done;
- writing (and presenting, if so decided by the PI) academic papers and materials for publication; and
- documenting in a sufficiently-detailed manner so as to share what you've learned.
You will be expected to participate in software development; this includes
- learning how to use third party tools, including editors, IDEs, databases, business process modeling software, programming languages, and libraries;
- evaluating competing software systems for suitability;
- building infrastructure to enable application development;
- managing yourself, and managing fellow team members;
- engaging in product, project, and program management as needed;
- reading books and consuming such other media forms about software development;
- writing documentation such as user/developer-facing manuals, tutorials;
- delivering community support to users and customers;
- preparing presentations to represent the project before external audiences; and
- writing actual user-facing applications throughout the "full stack"
in each case, documenting, reporting, and/or presenting the same in clearly and efficiently, both internally and externally.
Researchers under the Programme should possess both mastery of their academic speciality, and demonstrated practical experience producing and supporting high-quality open-source code through community engagement across multiple media. Candidates must have Bachelor's, Master's, Ph.D., or alternative qualifications in computer science, software engineering, (symbolic) artificial intelligence, computational linguistics, or formal methods.
Fluency in multiple natural languages, and any experience with law, are a plus.
This specialist role requires skills in the following areas (one or two in each category is sufficient):
Open-Source Software & Open-Standard Infrastructure Engineering Internet ꞏ Unix ꞏ TCP/IP ꞏ HTTP(S) ꞏ XML ꞏ JSON ꞏ Schemas ꞏ YAML ꞏ Git ꞏ Github ꞏ Technical Writing ꞏ Developer Relations and Evangelism
Formalization of Contracts, Rules, and Business Logic: Akoma Ntoso ꞏ LegalRuleML ꞏ Petri Nets ꞏ Drools ꞏ iLog ꞏ OPA ꞏ Neota Logic ꞏ Constraint Handling Rules ꞏ ICAIL ꞏ JURIX ꞏ RuleML+RR
Semantic Web and Ontologies: RDFa ꞏ RDFt ꞏ RDF ꞏ RIF ꞏ SWRL ꞏ OWL ꞏ Cyc ꞏ KBpedia ꞏ SUMO ꞏ UFO ꞏ Protégé ꞏ or a commercial rules engine with ontology support
Model Driven Architecture: UML ꞏ BPMN ꞏ DMN ꞏ OCL ꞏ SBVR ꞏ Statecharts ꞏ Activity Diagrams ꞏ Parnas Tables Logics: Hoare
Logic ꞏ Kripke structures ꞏ Modal mu-calculus ꞏ Dynamic Logic ꞏ Deontic, Temporal, and Epistemic modals ꞏ Situation Calculus ꞏ Process Calculi ꞏ Multivalent logics ꞏ Defeasible Logic ꞏ BDI ꞏ Abductive Logic ꞏ F-Logic ꞏ Intuitionistic ꞏ Non-monotonic Logics ꞏ Skolemization ꞏ Well-Founded Semantics Good Old-Fashioned Artificial Intelligence and Semantic Web
Knowledge Representation and Reasoning: Lisp ꞏ Prolog ꞏ F-Logic ꞏ Drools ꞏ Oracle Policy Manager ꞏ Flora-2 ꞏ Datalog ꞏ Answer-Set Programming ꞏ Defeasible Logic ꞏ Expert Systems ꞏ DocAssemble ꞏ lps.doc.ic.ac.uk
Skills listed below this line are preferred but not required:
Theorem Proving and Strongly Typed Languages: CoQ ꞏ Agda ꞏ Idris ꞏ Lean Prover ꞏ F* ꞏ Advanced Haskell (Template etc) ꞏ OCaml ꞏ Elm ꞏ Isabelle/HOL ꞏ StrangeLoop ꞏ Pony
Model Checkers and SAT/SMT solvers: Z3 ꞏ PAT ꞏ SMTlib ꞏ NuSMV ꞏ SPIN ꞏ Sentient
IDEs, Language Workbenches, and PLT frameworks: K Framework ꞏ JetBrains MPS ꞏ IDEs (Vim, Emacs, Visual Studio Code, Atom, Sublime Text) ꞏ Language Server Protocol Rule Systems and Logics
Computational Linguistics – Natural Language Generation Computational Semantics (in Haskell, Prolog, or Lisp) ꞏ Montague Semantics ꞏ Grammatical Framework ꞏ Text planning ꞏ Surface realization Familiarity with: WordNet ꞏ FrameNet ꞏ SenseNet ꞏ Controlled Natural Languages ꞏ Attempto Controlled English ꞏ ACErules
Parsers, Compilers, and Programming Language Theory Lex/YACC ꞏ alex/happy ꞏ Bison ꞏ CFGs ꞏ EBNF ꞏ BNFC
Formal Verification and Specification Languages Some familiarity with any of: TLA+ ꞏ Alloy ꞏ IVy ꞏ B Method ꞏ Z notation ꞏ VDM++ ꞏ LTL ꞏ CTL
App Development Test-Driven Development ꞏ Continuous Integration ꞏ Software Project Management ꞏ Agile / XP / Scrum ꞏ Pair Programming ꞏ Literate Programming ꞏ Unit Testing ꞏ Amazon Web Services ꞏ Docker ꞏ Full-Stack Development
Singapore Management University