Theresa Swift will give an invited talk, this Thursday, on a powerful
system supporting many of the most impressive logic and reasoning
features developed by the group of Prof. Kifer at Stony Brook. It's
exactly for those applications that require assurance, and LLMs cannot
do. All are welcome!  

Thursday, Nov 21, 2024, 11 AM-12:20 PM on Zoom
https://stonybrook.zoom.us/j/94298584108?pwd=OXRLemROdk8xOHY3NXFRV2xjQmhVUT09

Multi-paradigm Logic Programming in the ErgoAI System

Speaker:
Theresa Swift, Johns Hopkins University Applied Physics Laboratory

Abstract:

ErgoAI (abbr. Ergo) is a high level, multi-paradigm logic programming
language and system developed by Coherent Knowledge Systems and Stony
Brook University as an enhanced successor to the popular object-based
Flora system.  Ergo is oriented towards scalable knowledge
representation and reasoning, and can exploit both structured
knowledge as well as knowledge derived from external sources such as
vector embeddings.  From the start, Ergo (and Flora before it) were
designed to exploit the well-founded semantics for reasoning in a
multi-paradigm environment, including object-based frame logic with
non-monotonic inheritance; higher order syntax in the style of HiLog;
the optional addition of argumentation theories to modules;
semantically clean transactional updates; and extensive use of dynamic
goal reordering for better logical behavior and performance.

Although Ergo programs are compiled into XSB Prolog and adopt many
Prolog features, Ergo is altogether a different language with a scope
of functionality equivalent to a major programming language.  At the
same time, Ergo is tightly integrated with both Prolog and Python, so
that data structures of Ergo are quickly translated to data structures
of these other languages when needed.  Ergo has been used in many
commercial and research applications for companies, governmental
agencies, and research labs, such as: industrial manufacturing, legal
and financial applications, healthcare, and defense.
 

Bio: 

Theresa Swift's research background includes over 80 publications on
artificial intelligence and computational logic in leading journals
and conferences.  Specific topics include probabilistic reasoning and
reasoning over T-norms, description logics, model checking and process
logics, stream-based reasoning, non-monotonic and abductive reasoning,
hybrid reasoning, logical updates, data standardization, termination
theory, and logic programming semantics and implementation.  For 25
years Dr. Swift has co-led development of the open-source XSB logic
system.  XSB has been widely used in research projects around the
world, and by a number of companies including Google, SRI and BBN.