A globally optimal algorithm for TTD-MDPs

Bhat, Sooraj; Roberts, David L.; Nelson, Mark J.; Isbell, Charles L.; Mateas, Michael

doi:10.1145/1329125.1329367

Cited by 7 publications

(3 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…MDPs were first used in the context of experience management by the Declarative Optimization-based Drama Management (DODM) system (Nelson et al 2006a;2006b), which built on the search-based drama management pioneered by the Oz Project (Bates 1992;Weyhrauch 1997). MDPs serve as the basis for targeted trajectory distribution markov decision processes (TTD-MDPs) (Roberts et al 2006;Bhat et al 2007;Cantino, Roberts, and Isbell 2007;Roberts, Cantino, and Isbell Jr. 2007;). Created to provide replayability in the context of experience management, TTD-MDPs model desired distributions over outcomes and trajectories instead of maximizing expected reward.…”

Section: Related Workmentioning

confidence: 99%

Narrative Mediation as Probabilistic Planning

Robertson

Young

2021

AIIDE

View full text Add to dashboard Cite

Narrative mediation is a plan-based process that converts a planning problem into a mediation tree that models an interactive story. Probabilistic planning relaxes the classical planning constraint of deterministic action outcomes by containing domain operators with nondeterministic effects. This paper shows narrative mediation can be viewed as a probabilistic planning problem where the outcomes of user actions are modeled with nondeterministic effects in the domain. The paper gives a process for automatically converting a deterministic mediation planning problem into a probabilistic interactive narrative planning problem.

show abstract

Section: Related Workmentioning

confidence: 99%

Narrative Mediation as Probabilistic Planning

Robertson

Young

2021

AIIDE

View full text Add to dashboard Cite

show abstract

“…TTD-MDPs are a statistical machine learning formalism that leverage probabilistic reasoning to create a "policy" specifying the relative probability of short term goals in relation to the authors overall criteria. They have been applied to the goal selection problem in DODM (Roberts et al 2006;Bhat et al 2007;Roberts et al 2007;Roberts 2010).…”

Section: Representation Of Authors' Criteriamentioning

confidence: 99%

Seven Design Challenges for Fully-realized Experience Management

Roberts

2011

AIIDE

View full text Add to dashboard Cite

Drama Managers, a specific type of the more general Experience Manager, have become a common subject of study in the interactive narrative literature. With a range of representational and computational approaches, authors have repeatedly developed techniques that enable computers to generate, reason about, and adapt narratives in an interactive virtual setting. In order to fully realize an experience manager, seven representational and computational problems need to be solved, generally on a case-by-case basis. In other words, the choice to use an Experience Manager is the choice to model the design as, and implement solutions to, seven inter-dependent design problems. We explicitly articulate those design problems and provide a number of examples of methods that both motivate the design problems as well as illustrate a range of approaches to solving them.

show abstract

“…Techniques have been developed for tightly coupling plot creation and character behavior in dialogue-oriented interactive stories [13], for monitoring users' actions to determine if they are threatening the plot and, if so, either accommodating the new development or intervening [11] [18]. Search-based approaches utilize an evaluation function encoding author aesthetics to guide narrative planning [15] [22], while reinforcement learning has been used to learn policies to guide narrative action selection [16], and Targeted Trajectory Distribution Markov Decision Processes (TTD-MDP) have been employed to enhance replayability [4] [19]. Other narrative planners have taken decision-theoretic approaches to narrative management [14][21].…”

Section: Introductionmentioning

confidence: 99%

Investigating director agents' decision making in interactive narrative

Lee

Mott

Lester

2010

Proceedings of the Intelligent Narrative Technologies III Workshop

View full text Add to dashboard Cite

Interactive narrative planning offers significant potential for creating engaging narrative experiences that are tailored to individual users. Orchestrating all of the events in a storyworld to create optimal user experiences calls for effective narrative decision-making. A key requirement of this endeavor is understanding the role that different knowledge sources play in narrative decision making. To investigate knowledge sources for interactive narrative, a corpus was collected in a Wizard-of-Oz (WOZ) study conducted with a narrative-centered learning environment. With narrative planning and natural language dialogue functionalities provided by wizards, the data from the WOZ study offers insight into the knowledge sources involved in narrative decision making and suggests how these knowledge sources can be effectively utilized by a narrative planner to create engaging interactive narratives.

show abstract

A globally optimal algorithm for TTD-MDPs

Cited by 7 publications

References 15 publications

Narrative Mediation as Probabilistic Planning

Narrative Mediation as Probabilistic Planning

Seven Design Challenges for Fully-realized Experience Management

Investigating director agents' decision making in interactive narrative

Contact Info

Product

Resources

About