Use Interactive Visualization to Facilitate Game Designers to Understand and Design with Machine Learning
by: Jiachi Xie, M.S. Digital Media, Class of 2018
Machine learning has recently reached a level of maturity sufficient for being useful to make game design. Work exists on automatic analysis of game designs, which could lead to significant reduction of design and production time. The recent successes of implementing machine learning have led to a wave of interest from game designers. Obtaining an understanding of machine learning, however, is difficult. We propose a web-based interactive visualization tool to offer a direct manipulation solution to facilitate game designers to understand and design with machine learning. Evaluation of our application through a panel of experts indicated that the interactive visualization is well-suited for helping designers gaining an intuition of the machine learning process. While a successful initial step towards assisting level design, further research is needed to make machine learning more understandable and useful to game designers.
Touch-Enhanced Gesture Control Scheme
by: Chelsea Myers, M.S. Digital Media, Class of 2016
Touch-Enhanced Gesture (TEG) is an approach for improving gesture control by combining it with touch input to address its key shortcomings: speed, accuracy, and live mic syndrome. Our TEG prototype replaces crucial gestures with touchscreen commands and includes a virtual clutch to lessen live mic syndrome. To test our design, we implemented the new control scheme on a generic smartphone and compared its performance to the Myo, a wearable gesture control device. Results of a user study (n=30) show that our touch-enhanced gesture control scheme is faster and more accurate when executing selection commands. Our prototype’s virtual clutch was also easier for participants to use and yielded a lower percent error.
Design Patterns for Silent Player Characters
by: Bria Mears, M.S. Digital Media, Class of 2016
The silent player character (or silent protagonist) is a popular but vaguely understood type of player character in contemporary narrative-driven games. In this thesis, two types of silent player character types were created (expressive and projective characters) to explain the difference in the playing experience. Based on a survey of related games, a synthesized list of methods were built that designers can use to effectively communicate a SPC’s story. Motherhood, a short narrative experience featuring an expressive SPC, was created to test each method. Motherhood was tested by players for both story, and character interpretation. Results concluded that the design patterns developed within this study, used to communicate an expressive SPC, were largely successful in developing a pre-defined SPC who players were able to interpret as a character. All of the patterns were successful in impacting the interpretation of the SPC, as long as the information presented in the patterns is repetitive and clear.
Learnability Through Adaptive Discovery Tools in Voice User Interfaces
by: Anushay Furqan, M.S. Digital Media, Class of 2016
The invisible nature of VUIs has been attributed to challenging discoverability of VUIs. When discoverability is challenging, learnability can be compromised. Some 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 create DiscoverCal, a calendar application designed using an adaptive discovery tool to improve learnability in VUIs. DiscoverCal is created using Api.ai enabled wall mounted display at home. We identify characteristics of discovery tools created by researchers and extend their work by designing a system that adapts based on contextual relevance and user performance, in order to extend learnability beyond initial use. We find that an adaptive approach is slightly more favorable for learnability. However, further iterations to our design of adaptivity are necessary.
Understanding Inner Perspective Through the Little Match Girl
by: Demi Barzana, M.S. Digital Media Candidate
The little match girl takes the original Hans Christian Anderson story and creates a game that allows the player to venture through a dark city on a snowy night and not only help progress the little girl through the story but see the world through her eyes. This game support the research topic of how using the mechanic of switching between 3rd and 1st person views in a side-scroller game can offer more insight into a game character’s inner perspective. The little match girl adds a new twist to the platformer game genre.
Agency Informing Techniques: Communicating Player Agency in Interactive Narratives
by: Timothy Day, M.S. Digital Media Candidate
Player agency is the satisfying power a player feels when they act in a game and see the results of those actions. Research on agency is typically concerned with a player’s theoretical agency—their actual level of autonomy. Research on the feeling of agency that a player perceives, a player’s perceived agency, is less explored. This thesis explores this topic by investigating techniques utilized in games and research that give a player a sense of agency. We also created a new technique used in a new game called Found where players are told the nature of the outcome relating to an in game choice before they make it.
Automated Narrative Information Extraction Using Non-Linear Natural Language Processing Pipelines
by: Josep Valls, Ph.D. in Computer Science, Class of 2018
Computational narrative is an emergent field of research at the intersection of traditional narratology, artificial intelligence, natural language processing and cognitive science. Computational narrative focuses on methods to algorithmically analyze, model, and generate narratives. Moreover, computational models of narrative have applications to tasks such as machine translation, summarization or information extraction in the contexts of reporting, education, and entertainment.
My research focuses on the problem of automatically acquiring structured narrative information from natural language. I have focused on character extraction and narrative role identification from a corpus of Slavic folktales. The long-term goal of my research is to enable interactive applications to use natural language as input, for example, by generating games from existing literature. I am interested in systems that can automatically extract narrative structures from natural language stories. I have looked into applying state-of-the-art NLP techniques to extract key narrative elements and reuse them in the context of story generation systems. I expect to be able to develop new frameworks for NLP and story generation that can be used by interactive narrative authors and computer game designers to provide a better experience for their audience. At the same time, I believe current NLP systems can be complemented with commonsense and narrative domain knowledge in order to improve the performance of core NLP tasks in the domains of storytelling and interactive fiction. I also expect to be able to learn about issues when applying off-the-shelf NLP systems to the narrative domains in order to be able to transfer to other specific text domains.
Understanding Metagaming Mechanics in Interactive Storytelling
by: Erica M. Kleinman, M.S. Digital Media, Class of 2016
This thesis involved the analysis of existing games that utilize a specific metagaming-based mechanic, rewind/redo, in order to develop a formal vocabulary. This vocabulary informed the design of three versions of an experimental game. This interactive narrative game, titled Rough Draft, was used in a formal research study that gathered empirical data through qualitative and quantitative results analysis as well as observations of gameplay practices. Significant findings include qualitative response trends that support our theoretical claims as well as results that provide additional insights into best design practices for interactive story games that utilize a rewind/redo mechanic.
Arab Gamers: An Identity Inclusivity Study
by: Bushra Alfaraj, M.S. Digital Media, Class of 2016
Despite the copious amount of video games that have been distributed around the world, gamers who come from minority or underrepresented backgrounds may not encounter many representations of their respective identities. When they do, these representations are often reduced to stereotypes.
Among these underrepresented and stereotypical representations are depictions of Arabs and Arab culture in mainstream video games, where the most common depiction being of Arabs as terrorists in shooter games. In addition to the limited representations of Arabs in video games, there is not enough empirical data on how Arab gamers feel about the current state of how video games portray their culture.
This thesis collects data from Arab gamers about how they make sense of ethnic representations of their culture in video games. We analyze the data through an intersectional lens to understand how identity complexity applies to the perceptions of individual participants in the study and how they tie in with other underrepresented identity groups.
Indemnity: An Activist Game Conveying Latino Immigration Motivations for Social Awareness
by: Caroline Guevara, M.S. Digital Media, Class of 2015
Indemnity is an activist game designed to inform audiences on the life threatening dangers young Central Americans face on a daily basis. Based on the theoretical framework of procedural rhetoric, Indemnity combines game mechanics and interactive storytelling to achieve our design goal. Results from a user study confirmed Indemnity was successful by increasing awareness and sympathy for undocumented Latinos who are currently facing hardships in their country. Furthermore, it was able to pique interest in participants to learn more about the social issue.
Feeling Factory: A Digital Game for Prosody Improvement in Children with ASD
by: Natalie Lyon, M.S. Digital Media, Class of 2015
Previous research has shown that children with Autism Spectrum Disorder (ASD) tend to struggle with correctly identifying and producing prosodic cues and may have a predilection for using computer and games. However, there are currently no digital games that target prosody production and perception. The design of the digital game Feeling Factory explores how to combine prosodic speech therapy techniques with game design techniques. The goal of the game is to improve emotional and grammatical, productive and receptive prosody in high-functioning children with ASD. Feeling Factory uses a two-player design in order to balance engagement via the digital game with contextual generalizability via in-person conversation. A feasibility study was conducted consisting of semi-structured interviews with a panel of experts and children with ASD to help determine the potential benefits of this design model. The study resulted in a high recommendation from both groups.
Fostering Understanding of ADHD Through Perspective-Based Video Games
by: Tom Goldman, M.S. Digital Media, Class of 2014
This project presents an approach for facilitating understanding of Attention Deficit Hyperactivity Disorder (ADHD) through the procedural rhetoric of our persuasive video game 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.
Avian: Game Design and Player Metrics for Player Modeling in Educational Games
by: Justin Patterson, M.S. Digital Media, Class of 2014
This project explores an approach for player modeling in our educational game called Avian. The game mechanics of Avian are explicitly designed to engage four player types in distinctive ways based on motivational theory and its influence on observable player activities. The player types of interest in our research are Explorers and Goal-Seekers, with additional attention on how players socialize in game spaces to form subcategories of these two types. This allows us to collect data on how players interact with the game, review how their interactions relate to self-assessed player type preferences, and gauge how effective the game functions as an educational tool.