Goudoulakis, E (2014) DIEGESIS A multi-agent Digital Interactive Storytelling framework using planning and re-planning techniques. Doctoral thesis, Liverpool John Moores University.

Text 157941_2014GoudoulakisPhD.pdf - Published Version Download (7MB)

Abstract

In recent years, the field of Digital Interactive Storytelling (DIS) has become very popular both in academic circles, as well as in the gaming industry, in which stories are becoming a unique selling point. Academic research on DIS focuses in the search for techniques that allow the creation of systems that can generate dynamically interesting stories which are not linear and can change dynamically at runtime as a consequence of a player’s actions, therefore leading to different story endings.To reach this goal, DIS systems usually employ Artificial Intelligence planning and re-planning algorithms as part of their solution. There is a lack of algorithms created specifically for DIS purposes since most DIS systems use generic algorithms, and they do not usually assess if and why a given algorithm is the best solution for their purposes. Additionally, there is no unified way (e.g. in the form of a selection of metrics) to evaluate such systems and algorithms.To address these issues and to provide new solutions to the DIS field, we performed a review of related DIS systems and algorithms, and based on the critical analysis of that work we designed and implemented a novel multi-agent DIS framework called DIEGESIS, which includes –among other novel aspects- two new DIS-focused planning and re-planning algorithms.To ensure that our framework and its algorithms have met the specifications we set, we created a large scale evaluation scenario which models the story of Troy, derived from Homer’s epic poem, “Iliad”, which we used to perform a number of evaluations based on metrics that we chose and we consider valuable for the DIS field. This collection of requirements and evaluations could be used in the future from other DIS systems as a unified test-bed for analysis and evaluation of such systems.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	digital interactive storytelling, artificial intelligence, planning, re-planning, multi-agent systems
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software
Divisions:	Computer Science & Mathematics
Date Deposited:	20 Oct 2016 13:22
Last Modified:	03 Sep 2021 23:27
DOI or ID number:	10.24377/LJMU.t.00004512
Supervisors:	El Rhalibi, Abdennour, Merabti, Madjid and Taleb-Bendiab, A.
URI:	https://researchonline.ljmu.ac.uk/id/eprint/4512

View Item