mediKanren: A System for Bio-medical Reasoning
We introduce \emph{mediKanren}, a combination of miniKanren, a database describing relationships between medical concepts, and a graphical user interface (GUI) to simplify data exploration and common queries. All features of the faster-miniKanren implementation are available for queries, including typical constructs like run
, conde
, fresh
, various constraints, and (potentially recursive) user-defined relations. We provide the database as a set of miniKanren relations. To make queries fast, we represent the data backing these relations as specially formatted files on disk, with indexes for fast retrieval. We use the miniKanren project
syntax to write Racket code that interfaces with this representation. Performance is sufficient for the GUI to support low latency querying in the common case.
In this paper we provide an overview of mediKanren, give examples of queries in low-level and medium-level query languages, and give an example of how query results have resulted in treatment for a patient with a rare genetic disease.
Thu 27 AugDisplayed time zone: Eastern Time (US & Canada) change
15:30 - 17:10 | Afternoon SessionminiKanren at miniKanren Chair(s): Nada Amin Harvard University, Weixi Ma Indiana University | ||
15:30 20mTalk | mediKanren: A System for Bio-medical Reasoning miniKanren Michael Patton University of Alabama at Birmingham, Gregory Rosenblatt University of Alabama at Birmingham, USA, William E. Byrd University of Alabama at Birmingham, USA, Matthew Might University of Alabama at Birmingham | Harvard Medical School Link to publication | ||
15:50 20mTalk | Relational Synthesis for Pattern Matching miniKanren Dmitrii Kosarev JetBrains Research, Saint Petersburg State University, Dmitri Boulytchev St. Petersburg State University, St. Petersburg, Russia Link to publication | ||
16:10 20mTalk | Some Novel miniKanren Synthesis Tasks miniKanren Link to publication | ||
16:30 20mTalk | A Relational Interpreter for Synthesizing JavaScript miniKanren Artem Chirkov University of Toronto Mississauga, Gregory Rosenblatt University of Alabama at Birmingham, USA, Matthew Might University of Alabama at Birmingham | Harvard Medical School, Lisa Zhang University of Toronto Mississauga Link to publication | ||
16:50 20mTalk | dxo: A System for Relational Algebra and Differentiation miniKanren Link to publication |