JudgeD: a probabilistic datalog with dependencies

B. Wanders; Maurice van Keulen; Jan Flokstra

JudgeD: a probabilistic datalog with dependencies

Databases (Former)

Research output: Contribution to conference › Paper › peer-review

4 Citations (Scopus)

207 Downloads (Pure)

Abstract

We present JudgeD, a probabilistic datalog. A JudgeD program defines a distribution over a set of traditional datalog programs by attaching logical sentences to clauses to implicitly specify traditional data programs. Through the logical sentences, JudgeD provides a novel method for the expression of complex dependencies between both rules and facts. JudgeD is implemented as a proof-of-concept in the language Python. The implementation allows connection to external data sources, and features both a Monte Carlo probability approximation as well as an exact solver supported by BDDs. Several directions for future work are discussed and the implementation is released under the MIT license.

Original language	English
Pages	-
Number of pages	6
Publication status	Published - 13 Feb 2016
Event	Workshop on Declarative Learning Based Programming, DeLBP 2016 - Phoenix, United States Duration: 13 Feb 2016 → 13 Feb 2016

Workshop

Workshop	Workshop on Declarative Learning Based Programming, DeLBP 2016
Abbreviated title	DeLBP 2016
Country/Territory	United States
City	Phoenix
Period	13/02/16 → 13/02/16
Other	In conjunction with the Thirtieth AAAI Conference on Artificial Intelligence (AAAI-16)

Keywords

probabilistic Datalog
EWI-26661
METIS-315143
IR-98966
Datalog

Access to Document

delbp2016datalogFinal published version, 182 KB

http://www.aaai.org/ocs/index.php/WS/AAAIW16/paper/view/12564

Cite this

@conference{17a4188579eb4d52bd3f3b4cdbe52189,

title = "JudgeD: a probabilistic datalog with dependencies",

abstract = "We present JudgeD, a probabilistic datalog. A JudgeD program defines a distribution over a set of traditional datalog programs by attaching logical sentences to clauses to implicitly specify traditional data programs. Through the logical sentences, JudgeD provides a novel method for the expression of complex dependencies between both rules and facts. JudgeD is implemented as a proof-of-concept in the language Python. The implementation allows connection to external data sources, and features both a Monte Carlo probability approximation as well as an exact solver supported by BDDs. Several directions for future work are discussed and the implementation is released under the MIT license.",

keywords = "probabilistic Datalog, EWI-26661, METIS-315143, IR-98966, Datalog",

author = "B. Wanders and \{van Keulen\}, Maurice and Jan Flokstra",

year = "2016",

month = feb,

day = "13",

language = "English",

pages = "--",

note = "Workshop on Declarative Learning Based Programming, DeLBP 2016, DeLBP 2016 ; Conference date: 13-02-2016 Through 13-02-2016",

}

TY - CONF

T1 - JudgeD: a probabilistic datalog with dependencies

AU - Wanders, B.

AU - van Keulen, Maurice

AU - Flokstra, Jan

PY - 2016/2/13

Y1 - 2016/2/13

N2 - We present JudgeD, a probabilistic datalog. A JudgeD program defines a distribution over a set of traditional datalog programs by attaching logical sentences to clauses to implicitly specify traditional data programs. Through the logical sentences, JudgeD provides a novel method for the expression of complex dependencies between both rules and facts. JudgeD is implemented as a proof-of-concept in the language Python. The implementation allows connection to external data sources, and features both a Monte Carlo probability approximation as well as an exact solver supported by BDDs. Several directions for future work are discussed and the implementation is released under the MIT license.

AB - We present JudgeD, a probabilistic datalog. A JudgeD program defines a distribution over a set of traditional datalog programs by attaching logical sentences to clauses to implicitly specify traditional data programs. Through the logical sentences, JudgeD provides a novel method for the expression of complex dependencies between both rules and facts. JudgeD is implemented as a proof-of-concept in the language Python. The implementation allows connection to external data sources, and features both a Monte Carlo probability approximation as well as an exact solver supported by BDDs. Several directions for future work are discussed and the implementation is released under the MIT license.

KW - probabilistic Datalog

KW - EWI-26661

KW - METIS-315143

KW - IR-98966

KW - Datalog

M3 - Paper

SP - -

T2 - Workshop on Declarative Learning Based Programming, DeLBP 2016

Y2 - 13 February 2016 through 13 February 2016

ER -

JudgeD: a probabilistic datalog with dependencies

Abstract

Workshop

Keywords

Access to Document

Fingerprint

Cite this