Across the board, multi-source input is significantly better than single-source input in terms of accuracy. MSBART-F performs best on every measurement except Rogue-1 on WKH-R. 

Conclusion: The MSBART-F model (the model that includes input from multiple sources and encodes for relevance, abstraction, and consensus) generates the highest quality subtasks when compared to state-of-the-art variants.

Results: Research Question 1 (Learning to Decompose and Organize Complex Tasks)

MSBART-F infers dependencies with higher accuracy than the comparisons (prediction accuracy: 0.779, which is not fantastic but is pretty good). 

Results: Research Question 2 (Learning to Decompose and Organize Complex Tasks)

The reason the proposed model, MSBART-F, was as accurate as it was is that it encoded relevance, abstraction, and consensus .

Reasons for MSBART-F's Accuracy

1. Can we accurately and automatically generate subtasks given an input complex task?
2. Can we correctly identify the temporal dependencies between subtasks?

University of Rochester

Some researchers have used graph theory to create an end-to-end pipeline that creates a dependency graph from unstructured text. It represents subtasks and their time dependencies on each other. It is a way to automate the decomposition process.

Using Graph Theory to Decompose Tasks

Zhang, Y., Jauhar, S. K., Kiseleva, J., White, R., &amp; Roth, D. (2021). Learning to Decompose and Organize Complex Tasks. ACL Anthology. https://www.aclweb.org/anthology/2021.naacl-main.217/. 

Learning to Decompose and Organize Complex Tasks

Research Questions (Learning to Decompose and Organize Complex Tasks)

1. Task
2. Complex Task
3. Subtask
4. Temporal Dependencies

Terms Related to Modeling Tasks

The model is being treated as a graph induction problem, i.e. how to make a graph given the input. In graph theory, there are two main parts of a graph: nodes (points) and edges (what connects them). Here, nodes represent subtasks. The edges between the nodes represent temporal dependencies, which illustrate which tasks must be completed before another one is actionable. For example, you must create a guest list before you can send out invitations for a party. The graph is directed, which means it functions more like a flow chart instead of a plot of relations. The graph is induced from unstructured text found online, so it is a general and scalable approach.

Introduction to Graph Theory and How It Relates to Task Decomposition

These equations formalize "make a flow chart" so that the process can be automated.

Learn Before

Related

Learn After