Jichen Zhu

Towards story-based content generation: From plot-points to maps

Some computer game genres require meaningful stories and complex worlds in order to successfully engage players. In this paper we look at a procedural approach to story-based map generation focusing on the tight relationship between stories and the virtual worlds where those stories will unfold. Our long term goal is to develop procedural content generation techniques that can produce maps supporting multiple stories. We present an approach that takes, as input, the specification of a story space as a collection of plot points. Causal relations between these plot points and spatial relationships between locations define different story and spatial structures. Our system generates multiple configurations of a map, determines the stories that are actually supported in each map, and evaluates their quality, in order to find maps that support high quality stories from a storytelling perspective.

On the role of domain knowledge in analogy-based story generation

Computational narrative is a complex and interesting domain for exploring AI techniques that algorithmically analyze, understand, and most importantly, generate stories. This paper studies the importance of domain knowledge in story generation, and particularly in analogy-based story generation (ASG). Based on the construct of knowledge container in case-based reasoning, we present a theoretical framework for incorporating domain knowledge in ASG. We complement the framework with empirical results in our existing system Riu.

Story representation in analogy-based story generation in Riu

Computational analogy offers a promising direction to algorithmically generating stories, a key challenge in computational narrative. Since analogy methods are very sensitive to the story representation being used, this paper focuses on story representation for analogy-based story generation. Specifically, we analyze existing story representation formalisms and propose a new approach based on the cognitive semantics theory of force dynamics. Finally, we present the results of our analogy-based interactive narrative system, Riu, to illustrate the utility of our proposal.

Shall I Compare Thee to Another Story?—An Empirical Study of Analogy-Based Story Generation

Despite their use in traditional storytelling, analogy-based narrative devices have not been sufficiently explored in computational narrative. In this paper, we present our analogy-based story generation (ASG) approach in the Riu system, focusing on analogical retrieval and projection. We report on an empirical user evaluation about Rius capability to retrieve and generate short noninteractive stories using the story analogies through mapping (SAM) algorithm. This work provides the foundation for exploration of ASG in more complex and interactive computational narrative works.

Little Newton: an educational physics game

Little Newton is a 3D defense game in which the player learns about basic physics concepts by controlling physical attributes of projectiles. The mechanics of the game require the player to learn the basics of parabolic arcs, and friction in order to make use of the projectiles. Educational and learning theories are applied to the design in order to increase the ability of the player to learn how to play the game itself and therefore learn physics concepts in the process, balancing entertainment and educational content.

Back-Leading through character status in interactive storytelling

A key challenge in computer-based interactive narrative is the conflict between user agency and authorial control of the story quality. Valuable lessons can be learned from improvisational and especially interactive theatre, where various narrative and interactive strategies have been developed to engage users in the process of co-creating the story. In this paper, we focus on the use of character status and status shifts. Specifically, we present and illustrate a computational model of status shifts based on the cognitive semantics theory of force dynamics.

Learnability through Adaptive Discovery Tools in Voice User Interfaces

The invisible nature of VUIs has been attributed to challenging discoverability with VUIs. Low discoverability often leads to learnability issues. Researchers have designed visual tools for VUIs to help users learn as they go. However, few have used adaptation to ensure that learnability with the help of these tools extends beyond initial use. We designed DiscoverCal, a calendar application designed using adaptive discovery tools to improve learnability in VUIs. In this paper, we identify key characteristics of existing discovery tools. We present our design of a VUI that adapts based on contextual relevance and user performance in order to extend learnability beyond initial use. We briefly discuss our user study design.

Towards balancing learner autonomy and pedagogical process in educational games

We present the preliminary work in the TAEMILE project, which aims to co-regulate the learning process in educational games by automatically balancing learners autonomy and the pedagogical processes intended by educators. We focus on our design rationale and the initial results from our user study.

Using video games to facilitate understanding of attention deficit hyperactivity disorder: a feasibility study

This paper presents an approach for facilitating understanding of Attention Deficit Hyperactivity Disorder (ADHD) through the procedural rhetoric of our persuasive video game { em Drawn to Distraction}. Different from realistic simulations, our game is designed to convey a message about the disorder primarily through game mechanics. To test the feasibility of this approach, we conducted a series of studies involving caregivers of ADHD-affected children and the general public. The results, especially in Experiment 3, show promising trends on the feasibility of using persuasive games to promote understanding of psychological disorders.

Interactive non-fiction: towards a new approach for storytelling in digital journalism

The development of digital journalism, an important area of storytelling, has focused primarily on incorporating interactive multimedia components to traditional linear news stories. In this paper, we propose a new approach that places central focus on text, the core of journalism, based on insights from fictional interactive digital storytelling. Drawing upon the convention of interactive fiction, we present our preliminary work on interactive non-fiction.