The most probable explanation for probabilistic logic programs with annotated disjunctions

Dimitar Shterionov; Joris Renkens; Jonas Vlasselaer; Angelika Kimmig; Wannes Meert; Gerda Janssens

doi:10.1007/978-3-319-23708-4_10

The most probable explanation for probabilistic logic programs with annotated disjunctions

Dimitar Shterionov^*, Joris Renkens, Jonas Vlasselaer, Angelika Kimmig, Wannes Meert, Gerda Janssens

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

10 Citations (Scopus)

Abstract

Probabilistic logic languages, such as ProbLog and CP-logic, are probabilistic generalizations of logic programming that allow one to model probability distributions over complex, structured domains. Their key probabilistic constructs are probabilistic facts and annotated disjunctions to represent binary and mutli-valued random variables, respectively. ProbLog allows the use of annotated disjunctions by translating them into probabilistic facts and rules. This encoding is tailored towards the task of computing the marginal probability of a query given evidence (MARG), but is not correct for the task of finding the most probable explanation (MPE) with important applications e.g., diagnostics and scheduling. In this work, we propose a new encoding of annotated disjunctions which allows correct MARG and MPE. We explore from both theoretical and experimental perspective the trade-off between the encoding suitable only for MARG inference and the newly proposed (general) approach.

Original language	English
Title of host publication	Inductive Logic Programming - 24th International Conference, ILP 2014, Revised Selected Papers
Editors	Jesse Davis, Jan Ramon
Publisher	Springer Verlag
Pages	139-153
Number of pages	15
ISBN (Print)	9783319237077
DOIs	https://doi.org/10.1007/978-3-319-23708-4_10
Publication status	Published - 2015
Externally published	Yes
Event	24th International Conference on Inductive Logic Programming, ILP 2014 - Nancy, France Duration: 14 Sept 2014 → 16 Sept 2014

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9046
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	24th International Conference on Inductive Logic Programming, ILP 2014
Country/Territory	France
City	Nancy
Period	14/09/14 → 16/09/14

Keywords

Logic programs with annotated disjunctions
Most Probable Explanation
Probabilistic logic programming
ProbLog
Statistical relational learning

Access to Document

10.1007/978-3-319-23708-4_10

Cite this

Shterionov, D., Renkens, J., Vlasselaer, J., Kimmig, A., Meert, W., & Janssens, G. (2015). The most probable explanation for probabilistic logic programs with annotated disjunctions. In J. Davis, & J. Ramon (Eds.), Inductive Logic Programming - 24th International Conference, ILP 2014, Revised Selected Papers (pp. 139-153). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9046). Springer Verlag. https://doi.org/10.1007/978-3-319-23708-4_10

Shterionov, Dimitar ; Renkens, Joris ; Vlasselaer, Jonas et al. / The most probable explanation for probabilistic logic programs with annotated disjunctions. Inductive Logic Programming - 24th International Conference, ILP 2014, Revised Selected Papers. editor / Jesse Davis ; Jan Ramon. Springer Verlag, 2015. pp. 139-153 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{84b457d56ae845d883f85eff5cd0b259,

title = "The most probable explanation for probabilistic logic programs with annotated disjunctions",

abstract = "Probabilistic logic languages, such as ProbLog and CP-logic, are probabilistic generalizations of logic programming that allow one to model probability distributions over complex, structured domains. Their key probabilistic constructs are probabilistic facts and annotated disjunctions to represent binary and mutli-valued random variables, respectively. ProbLog allows the use of annotated disjunctions by translating them into probabilistic facts and rules. This encoding is tailored towards the task of computing the marginal probability of a query given evidence (MARG), but is not correct for the task of finding the most probable explanation (MPE) with important applications e.g., diagnostics and scheduling. In this work, we propose a new encoding of annotated disjunctions which allows correct MARG and MPE. We explore from both theoretical and experimental perspective the trade-off between the encoding suitable only for MARG inference and the newly proposed (general) approach.",

keywords = "Logic programs with annotated disjunctions, Most Probable Explanation, Probabilistic logic programming, ProbLog, Statistical relational learning",

author = "Dimitar Shterionov and Joris Renkens and Jonas Vlasselaer and Angelika Kimmig and Wannes Meert and Gerda Janssens",

year = "2015",

doi = "10.1007/978-3-319-23708-4_10",

language = "English",

isbn = "9783319237077",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "139--153",

editor = "Jesse Davis and Jan Ramon",

booktitle = "Inductive Logic Programming - 24th International Conference, ILP 2014, Revised Selected Papers",

address = "Germany",

note = "24th International Conference on Inductive Logic Programming, ILP 2014 ; Conference date: 14-09-2014 Through 16-09-2014",

}

Shterionov, D, Renkens, J, Vlasselaer, J, Kimmig, A, Meert, W & Janssens, G 2015, The most probable explanation for probabilistic logic programs with annotated disjunctions. in J Davis & J Ramon (eds), Inductive Logic Programming - 24th International Conference, ILP 2014, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9046, Springer Verlag, pp. 139-153, 24th International Conference on Inductive Logic Programming, ILP 2014, Nancy, France, 14/09/14. https://doi.org/10.1007/978-3-319-23708-4_10

The most probable explanation for probabilistic logic programs with annotated disjunctions. / Shterionov, Dimitar; Renkens, Joris; Vlasselaer, Jonas et al.
Inductive Logic Programming - 24th International Conference, ILP 2014, Revised Selected Papers. ed. / Jesse Davis; Jan Ramon. Springer Verlag, 2015. p. 139-153 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9046).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - The most probable explanation for probabilistic logic programs with annotated disjunctions

AU - Shterionov, Dimitar

AU - Renkens, Joris

AU - Vlasselaer, Jonas

AU - Kimmig, Angelika

AU - Meert, Wannes

AU - Janssens, Gerda

PY - 2015

Y1 - 2015

N2 - Probabilistic logic languages, such as ProbLog and CP-logic, are probabilistic generalizations of logic programming that allow one to model probability distributions over complex, structured domains. Their key probabilistic constructs are probabilistic facts and annotated disjunctions to represent binary and mutli-valued random variables, respectively. ProbLog allows the use of annotated disjunctions by translating them into probabilistic facts and rules. This encoding is tailored towards the task of computing the marginal probability of a query given evidence (MARG), but is not correct for the task of finding the most probable explanation (MPE) with important applications e.g., diagnostics and scheduling. In this work, we propose a new encoding of annotated disjunctions which allows correct MARG and MPE. We explore from both theoretical and experimental perspective the trade-off between the encoding suitable only for MARG inference and the newly proposed (general) approach.

AB - Probabilistic logic languages, such as ProbLog and CP-logic, are probabilistic generalizations of logic programming that allow one to model probability distributions over complex, structured domains. Their key probabilistic constructs are probabilistic facts and annotated disjunctions to represent binary and mutli-valued random variables, respectively. ProbLog allows the use of annotated disjunctions by translating them into probabilistic facts and rules. This encoding is tailored towards the task of computing the marginal probability of a query given evidence (MARG), but is not correct for the task of finding the most probable explanation (MPE) with important applications e.g., diagnostics and scheduling. In this work, we propose a new encoding of annotated disjunctions which allows correct MARG and MPE. We explore from both theoretical and experimental perspective the trade-off between the encoding suitable only for MARG inference and the newly proposed (general) approach.

KW - Logic programs with annotated disjunctions

KW - Most Probable Explanation

KW - Probabilistic logic programming

KW - ProbLog

KW - Statistical relational learning

UR - http://www.scopus.com/inward/record.url?scp=84954088095&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-23708-4_10

DO - 10.1007/978-3-319-23708-4_10

M3 - Conference contribution

AN - SCOPUS:84954088095

SN - 9783319237077

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 139

EP - 153

BT - Inductive Logic Programming - 24th International Conference, ILP 2014, Revised Selected Papers

A2 - Davis, Jesse

A2 - Ramon, Jan

PB - Springer Verlag

T2 - 24th International Conference on Inductive Logic Programming, ILP 2014

Y2 - 14 September 2014 through 16 September 2014

ER -

Shterionov D, Renkens J, Vlasselaer J, Kimmig A, Meert W, Janssens G. The most probable explanation for probabilistic logic programs with annotated disjunctions. In Davis J, Ramon J, editors, Inductive Logic Programming - 24th International Conference, ILP 2014, Revised Selected Papers. Springer Verlag. 2015. p. 139-153. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-23708-4_10

The most probable explanation for probabilistic logic programs with annotated disjunctions

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this