Skip to content

Task Trees: hierarchical knowledge representation for task recognition and event selection

Design · All designs
Evidence · no claims cited

Description

Task Trees are the knowledge structure used by both the Task Model Reasoner and the Narrative Manager in the Bio-safety Lab. A tree's root represents the whole task, intermediate nodes are subtasks, and leaves are Task-Level Events; sibling nodes relate through AND, OR and SEQ conditions, and leaf nodes marked as errors let the system recognize plausible but incorrect actions. The article states "A Task Tree is a hierarchical model which has a root node representing the whole task to perform on the top of the tree and child nodes being the subtasks in which it can be divided". Narrative Manager trees additionally mark system nodes with Balance, Impact and Closure parameters.

Design Implications

Context

Requirements

  • The spill-clearing task tree contains more than 250 nodes, so building one requires detailed protocol knowledge and expert annotation of error nodes

Constraints

  • The article contrasts this with Death Kitchen's Danger Matrix, which allowed only one constraint (partial order) for event selection

Target Learners

  • Bio-safety laboratory workers in training
  • Medical students

Learning Goals

  • Recognition and correction of user errors during protocol execution
  • Real-time monitoring of multi-step laboratory procedures

Claims

Examples

-

Key Sources

  • Alvarez, N., Sanchez-Ruiz, A., Cavazza, M., Shigematsu, M., & Prendinger, H. (2014). Narrative Balance Management in an Intelligent Biosafety Training Application for Improving User Performance. Int J Artif Intell Educ 25, 35–59. https://doi.org/10.1007/s40593-014-0022-z
Page metadata
Field Value
type design
id task-trees-knowledge-representation
title Task Trees: hierarchical knowledge representation for task recognition and event selection
description Task Trees are the knowledge structure used by both the Task Model Reasoner and the Narrative Manager in the Bio-safety Lab.
status review
generated by process:settle-candidates on 2026-10-27
trust unverified — no one has confirmed this page's content
sources 1 — alvarez-2014

Frontmatter is the page's machine-readable record — see the Schema & Guide.