Skip to content

The Event Calculus in Probabilistic Logic Programs with Annotated Disjunctions

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Original languageEnglish
Title of host publicationAAMAS '17: Proceedings of the 2017 International Conference on Autonomous Agents & Multiagent Systems
Subtitle of host publicationMay 8–12, 2017, São Paulo, Brazil
Publisher or commissioning bodyACM
Number of pages9
StateAccepted/In press - 24 Jan 2017

Abstract

We propose a new probabilistic extension to the event calculus using the probabilistic logic programming (PLP) language ProbLog, and a language construct called the annotated disjunction. This is the first extension of the event calculus capable of handling numerous sources of uncertainty (e.g. from primitive event observations and from composite event definitions). It is also the first extension capable of handling multiple sources of event observations (e.g. in multi-sensor environments). We describe characteristics of this new extension (e.g. rationality of conclusions), and prove some important properties (e.g. validity in ProbLog). Our extension is directly implementable in ProbLog, and we successfully apply it to the problem of activity recognition under uncertainty in an event detection data set obtained from vision analytics of bus surveillance video.

Research areas

  • The event calculus, event reasoning, probabilistic logic programming, ProbLog, annotated disjunction

Documents

Documents

  • Full-text PDF (accepted author manuscript)

    Accepted author manuscript, 329 KB, PDF-document

    Embargo ends: 1/01/99

    Request copy

View research connections

Related faculties, schools or groups