Lifted Temporal Most Probable Explanation

Marcel Gehrke; Tanya Braun; Ralf Möller

doi:10.1007/978-3-030-23182-8_6

Lifted Temporal Most Probable Explanation

Marcel Gehrke^*, Tanya Braun, Ralf Möller

^*Corresponding author for this work

Institute of Information Systems

Abstract

The lifted dynamic junction tree algorithm (LDJT) answers filtering and prediction queries efficiently for temporal probabilistic relational models by building and then reusing a first-order cluster representation of a knowledge base for multiple queries and time steps. Another type of query asks for a most probable explanation (MPE) for given events. Specifically, this paper contributes (i) LDJT^mpe to efficiently solve the temporal MPE problem for temporal probabilistic relational models and (ii) a combination of LDJT and LDJT^mpe to efficiently answer assignment queries for a given number of time steps.

Original language	English
Title of host publication	ICCS 2019: Graph-Based Representation and Reasoning
Editors	Dominik Endres, Mehwish Alam, Diana Şotropa
Number of pages	14
Volume	11530 LNAI
Publisher	Springer, Cham
Publication date	19.06.2019
Pages	72-85
ISBN (Print)	978-3-030-23181-1
ISBN (Electronic)	978-3-030-23182-8
DOIs	https://doi.org/10.1007/978-3-030-23182-8_6
Publication status	Published - 19.06.2019
Event	24th International Conference on Conceptual Structures - Marburg, Germany Duration: 01.07.2019 → 04.07.2019 Conference number: 227759

Research Areas and Centers

Centers: Center for Artificial Intelligence Luebeck (ZKIL)
Research Area: Intelligent Systems

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/978-3-030-23182-8_6

Cite this

@inproceedings{0eefeb47a2384d94b9a35c5d198dc928,

title = "Lifted Temporal Most Probable Explanation",

abstract = "The lifted dynamic junction tree algorithm (LDJT) answers filtering and prediction queries efficiently for temporal probabilistic relational models by building and then reusing a first-order cluster representation of a knowledge base for multiple queries and time steps. Another type of query asks for a most probable explanation (MPE) for given events. Specifically, this paper contributes (i) LDJTmpe to efficiently solve the temporal MPE problem for temporal probabilistic relational models and (ii) a combination of LDJT and LDJTmpe to efficiently answer assignment queries for a given number of time steps.",

author = "Marcel Gehrke and Tanya Braun and Ralf M{\"o}ller",

note = "Funding Information: This research originated from the Big Data project being part of Joint Lab 1, funded by Cisco Systems Germany, at the centre COPICOH, University of L{\"u}beck. Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2019. Copyright: Copyright 2019 Elsevier B.V., All rights reserved.; 24th International Conference on Conceptual Structures , ICCS 2019 ; Conference date: 01-07-2019 Through 04-07-2019",

year = "2019",

month = jun,

day = "19",

doi = "10.1007/978-3-030-23182-8_6",

language = "English",

isbn = "978-3-030-23181-1",

volume = "11530 LNAI",

series = "Lecture Notes in Computer Science",

publisher = "Springer, Cham",

pages = "72--85",

editor = "Endres, {Dominik } and Alam, {Mehwish } and {\c S}otropa, {Diana }",

booktitle = "ICCS 2019: Graph-Based Representation and Reasoning",

}

Gehrke, M, Braun, T & Möller, R 2019, Lifted Temporal Most Probable Explanation. in D Endres, M Alam & D Şotropa (eds), ICCS 2019: Graph-Based Representation and Reasoning. vol. 11530 LNAI, Lecture Notes in Computer Science, vol. 11530 LNAI, Springer, Cham, pp. 72-85, 24th International Conference on Conceptual Structures , Marburg, Hesse, Germany, 01.07.19. https://doi.org/10.1007/978-3-030-23182-8_6

Lifted Temporal Most Probable Explanation. / Gehrke, Marcel; Braun, Tanya; Möller, Ralf.
ICCS 2019: Graph-Based Representation and Reasoning. ed. / Dominik Endres; Mehwish Alam; Diana Şotropa. Vol. 11530 LNAI Springer, Cham, 2019. p. 72-85 (Lecture Notes in Computer Science; Vol. 11530 LNAI).

Research output: Chapters in Books/Reports/Conference Proceedings › Conference contribution › peer-review

TY - GEN

T1 - Lifted Temporal Most Probable Explanation

AU - Gehrke, Marcel

AU - Braun, Tanya

AU - Möller, Ralf

N1 - Conference code: 227759

PY - 2019/6/19

Y1 - 2019/6/19

N2 - The lifted dynamic junction tree algorithm (LDJT) answers filtering and prediction queries efficiently for temporal probabilistic relational models by building and then reusing a first-order cluster representation of a knowledge base for multiple queries and time steps. Another type of query asks for a most probable explanation (MPE) for given events. Specifically, this paper contributes (i) LDJTmpe to efficiently solve the temporal MPE problem for temporal probabilistic relational models and (ii) a combination of LDJT and LDJTmpe to efficiently answer assignment queries for a given number of time steps.

AB - The lifted dynamic junction tree algorithm (LDJT) answers filtering and prediction queries efficiently for temporal probabilistic relational models by building and then reusing a first-order cluster representation of a knowledge base for multiple queries and time steps. Another type of query asks for a most probable explanation (MPE) for given events. Specifically, this paper contributes (i) LDJTmpe to efficiently solve the temporal MPE problem for temporal probabilistic relational models and (ii) a combination of LDJT and LDJTmpe to efficiently answer assignment queries for a given number of time steps.

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-85068619556&origin=inward&txGid=fcf908f6f7903476076d7e2e3d815a6a

U2 - 10.1007/978-3-030-23182-8_6

DO - 10.1007/978-3-030-23182-8_6

M3 - Conference contribution

SN - 978-3-030-23181-1

VL - 11530 LNAI

T3 - Lecture Notes in Computer Science

SP - 72

EP - 85

BT - ICCS 2019: Graph-Based Representation and Reasoning

A2 - Endres, Dominik

A2 - Alam, Mehwish

A2 - Şotropa, Diana

PB - Springer, Cham

T2 - 24th International Conference on Conceptual Structures

Y2 - 1 July 2019 through 4 July 2019

ER -

Lifted Temporal Most Probable Explanation

Abstract

Research Areas and Centers

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this