=Paper=
{{Paper
|id=Vol-1875/paper18
|storemode=property
|title=Rulelog: Highly Expressive Semantic Rules with Scalable Deep Reasoning
|pdfUrl=https://ceur-ws.org/Vol-1875/paper18.pdf
|volume=Vol-1875
|authors=Benjamin Grosof,Michael Kifer,Paul Fodor
|dblpUrl=https://dblp.org/rec/conf/ruleml/GrosofKF17
}}
==Rulelog: Highly Expressive Semantic Rules with Scalable Deep Reasoning==
https://ceur-ws.org/Vol-1875/paper18.pdf
Rulelog: Highly Expressive Semantic Rules with
Scalable Deep Reasoning
Benjamin N. Grosof1 , Michael Kifer23 , and Paul Fodor23
1
Accenture, USA
2
Coherent Knowledge Systems, USA
3
Stony Brook University, New York, USA
michael.kifer, paul.fodor@coherentknowledge.com
Abstract. In this half-day tutorial, we cover the fundamental concepts, key tech-
nologies, emerging applications, recent progress, and outstanding research issues
in the area of Rulelog, a leading approach to fully semantic rule-based knowledge
representation and reasoning (KRR). Rulelog matches well many of the require-
ments of cognitive computing. It combines deep logical/probabilistic reasoning
tightly with natural language processing (NLP), and complements machine learn-
ing (ML). Rulelog interoperates and composes well with graph databases, re-
lational databases, spreadsheets, XML, and expressively simpler rule/ontology
systems and can orchestrate overall hybrid KRR. Developed mainly since 2005,
Rulelog is much more expressively powerful than the previous state-of-the-art
practical KRR approaches, yet is computationally affordable. It is fully seman-
tic and has capable efficient implementations that leverage methods from logic
programming and databases, including dependency-aware smart caching and a
dynamic compilation stack architecture.
Keywords: knowledge representation and reasoning, declarative logic programs, se-
mantic rules, cognitive computing, natural language processing
1 Rulelog in Ergo
Rulelog extends Datalog (database logic) with general classical-logic-like formulas, in-
cluding existentials and disjunctions, and strong capabilities for meta knowledge and
reasoning, including higher-order syntax, flexible defeasibility and probabilistic uncer-
tainty, and restraint bounded rationality that ensures worst-case polynomial time for
query answering4 . A large subset of Rulelog is in draft as an industry standard. An ex-
citing research frontier is that Rulelog can combine closely with NLP to both interpret
and generate English, including potentially for conversational NL interaction.
The most complete system today for Rulelog is Ergo Suite (Ergo) from Coher-
ent Knowledge5 . A subset of Rulelog is also implemented in the open-source system
Flora-26 (a.k.a. Ergo Lite) and an earlier SILK system from Vulcan. Using Ergo rule
4
Benjamin Grosof’s work done mainly while at Coherent Knowledge.
5
http://coherentknowledge.com
6
http://flora.sourceforge.net
2 Grosof et. al
inference engine, we will illustrate Rulelog’s applications in deep reasoning and repre-
senting complex knowledge such as policies, regulations/contracts, science, and termi-
nology mappings across a wide range of tasks and domains in business, government,
and academe. Examples include: legal/policy compliance, e.g., in financial services;
financial reporting/accounting; health care treatment guidance and insurance; educa-
tion/tutoring; security/confidentiality policies; and e-commerce marketing.
In this tutorial on Rulelog as it is implemented in Ergo, we will cover some unique
KRR features, such as:
– frame-based object-oriented frame syntax [5] and higher-order statements [3, 10]
for practical logical knowledge representation;
– rule identifiers and provenance, argumentation-based defeasible reasoning [8, 9];
– general quantification and general formulas [4];
– ErgoText, an integration of logic with controlled natural language phrases (sprin-
kled with variables and other syntactic elements), which are translated to logic sen-
tences;
– probabilistic reasoning and weighted uncertainty with restraint bounded rationality,
including distribution semantics [7], and evidential probability [6];
– external querying and virtual data stores;
– dynamically evolving knowledge [2] and hypothetical reasoning, including integrity
constraints and alarms;
– explanations that are fully detailed, interactively navigable, and presented in natural
language – understandable by those who are not expert in logic or programming [1].
Much of this tutorial will be dedicated to applications of Ergo, both horizontally
(e.g., policy-based decisions, info Integration, analytics, human-computer interaction
(HCI), search, business intelligence, risk management) and vertically (e-commerce and
marketing, financial services, personalized e-learning, security and defense, biomedical,
insurance, Internet of Things (IoT), social media sharing policies).
Finally, we will discuss open research topics in Ergo, such as, authoring rules start-
ing from NL, distributed reasoning, optimization of uncertainty reasoning, equality,
aggregates, integration with ASP, constraint solving, and classical logic, hypotheticals,
abduction and integration with ML.
The goal of the tutorial is for the audience to walk away with an understanding
of Rulelogs key innovative logical and inferencing concepts, its broad applicability, its
overall advantages and limitations, a sample of some specific application areas, and its
open research topics.
The intended audience for this tutorial is the rules and reasoning community (all
of the RuleML+RR audience) and the assumed background of the participants is only
the basics of first-order-logic and relational databases. Knowledge of declarative logic
programs, XML, RDF, and SPARQL will be helpful but not required.
References
1. Andersen, C., Benyo, B., Calejo, M., Dean, M., Fodor, P., Grosof, B.N., Kifer, M., Liang,
S., Swift, T.: Advanced knowledge base debugging for rulelog. In: Joint Proceedings of the
Rulelog: Highly Expressive Semantic Rules with Scalable Deep Reasoning 3
7th International Rule Challenge, the Special Track on Human Language Technology and
the 3rd RuleML Doctoral Consortium, Seattle, USA, July 11 -13, 2013 (2013), http://
ceur-ws.org/Vol-1004/paper8.pdf
2. Bonner, A., Kifer, M.: Transaction logic: Unifying declarative and procedural knowledge
(1993), manuscript
3. Chen, W., Kifer, M., Warren, D.: HiLog: A foundation for higher-order logic programming.
Journal of Logic Programming 15(3), 187–230 (February 1993)
4. Grosof, B.: Rapid text-based authoring of defeasible higher-order logic formulas, via tex-
tual logic and rulelog. In: Morgenstern, L., Stefaneas, P., Lvy, F., Wyner, A., Paschke, A.
(eds.) Theory, Practice, and Applications of Rules on the Web, Lecture Notes in Computer
Science, vol. 8035, pp. 2–11. Springer Berlin Heidelberg (2013), http://dx.doi.org/
10.1007/978-3-642-39617-5_2
5. Kifer, M., Lausen, G., Wu, J.: Logical foundations of object-oriented and frame-based lan-
guages. Journal of ACM 42, 741–843 (July 1995)
6. Kyburg, H., Teng, C.: Uncertain Inference. Uncertain Inference, Cambridge University Press
(2001)
7. Riguzzi, F., Swift, T.: Well-definedness and efficient inference for probabilistic logic pro-
gramming under the distribution semantics. Theory and Practice of Logic Programming
13(2), 279–302 (2013)
8. Wan, H., Grosof, B., Kifer, M., Fodor, P., Liang, S.: Logic programming with defaults and
argumentation theories. In: Int’l Conference on Logic Programming (July 2009)
9. Wan, H., Kifer, M., Grosof, B.: Defeasibility in answer set programs with defaults and argu-
mentation rules. Semantic Web Journal (2014)
10. Yang, G., Kifer, M.: Reasoning about anonymous resources and meta statements on the Se-
mantic Web. Journal on Data Semantics, LNCS 2800 1, 69–98 (September 2003)