Growing interest in eXplainable Artificial Intelligence (XAI) aims to make AI and machine learning more understandable human users. However, most existing work focuses on new algorithms, not usability, practical interpretability efficacy real In this vision paper, we propose a research area of for Designers (XAID), specifically game designers. By focusing specific user group, their needs tasks, human-centered approach facilitating designers co-create with AI/ML techniques through XAID. We...

Voice User Interfaces (VUIs) are growing in popularity. However, even the most current VUIs regularly cause frustration for their users. Very few studies exist on what people do to overcome VUI problems they encounter, or how can be designed aid when these occur. In this paper, we analyze empirical data users (n=12) interact with our calendar system, DiscoverCal, over three sessions. particular, identify main obstacle categories and types of tactics participants employ them. We analyzed...

Voice User Interfaces (VUIs) are increasing in popularity. However, their invisible nature with no or limited visuals makes it difficult for users to interact unfamiliar VUIs. We analyze the impact of user characteristics and preferences on how a VUI-based calendar, DiscoverCal. While recent VUI studies behavior through self-reported data, we extend this research by analyzing both usage data observe correlations between types. Results from our study (n=50) led four key findings: 1)...

Abstract In this paper, we present a new deep-learning disruption-prediction algorithm based on important findings from explorative data analysis which effectively allows knowledge transfer existing devices to ones, thereby predicting disruptions using very limited disruption the devices. The analysis, conducted via unsupervised clustering techniques confirms that time-sequence are much better separators of disruptive and non-disruptive behavior than instantaneous plasma-state data, with...

The advent of artificial intelligence (AI) and machine learning (ML) bring human-AI interaction to the forefront HCI research. This paper argues that games are an ideal domain for studying experimenting with how humans interact AI. Through a systematic survey neural network (n = 38), we identified dominant metaphors AI patterns in these games. In addition, applied existing guidelines further shed light on player-AI context AI-infused systems. Our core finding is as play can expand current...

Identifying explainable player strategies and decision patterns that give insights into behavior is one of the most difficult tasks for game analytics, yet yields great informative potential various purposes. Industrial stakeholders can capture experience infer issues or feedback on design, content, balancing - while players themselves might want to leverage this technique contrast their style play other players, fostering self-regulated learning. On top that, in case educational games,...

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

In the context of a learning game to teach parallel programming, we describe procedural content generation (PCG) approach that can be controlled generate programming puzzles involving desired set concepts, and size "difficulty". Our is based on grammars control puzzle structure, orthographic graph embedding techniques render it into two-dimensional grid for our game. The proposed PCG system designed work with player model in order provide personalized experiences. We present an evaluation...

Existing work on player modeling often assumes that the play style of players is static. However, our recent shows evidence regularly change their over time. In this paper we propose a novel framework to capture by using episodic information and sequential machine learning techniques. particular, experiment with different trace segmentation strategies for prediction. We evaluate new gameplay data gathered from game-based interactive environment. Our results show techniques incorporate...

This paper presents an approach for automatically identifying high-level narrative structure information, particularly character roles, from unannotated folk tales. We introduce a new representation called em action matrices to encode Propp's theory on role and their sphere of action. tested our in fully automated system (Voz) using corpus 10 tales.Our experimental evaluation shows that capture useful information identification, provides insight into the error introduced by individual steps,...

Abstract The ability to identify underlying disruption precursors is key avoidance. In this paper, we present an integrated deep learning (DL) based model that combines prediction with the identification of several like rotating modes, locked H-to-L back transitions and radiative collapses. first part our study demonstrates DL-based unstable event identifier trained on 160 manually labeled DIII-D shots can achieve, average, 84% rate various frequent events (like H-L transition, mode,...

Ongoing efforts to turn Machine Learning (ML) into a design material have encountered limited success. This paper examines the burgeoning area of AI art understand how artists incorporate ML in their creative work. Drawing upon related HCI theories, we investigate create ambiguity by analyzing nine artworks that use computer vision and image synthesis. Our analysis shows that, addition established types ambiguity, worked closely with process (dataset curation, model training, application)...

Concurrent and parallel programming (CPP) skills are increasingly important in today's world of hardware. However, the conceptual leap from deterministic sequential to CPP is notoriously challenging make. Our educational game Parallel designed support learning core concepts through a game-based approach, focusing on connection between gameplay CPP. Through 10-week user study (n 25) an undergraduate concurrent course, first empirical for game, our results show that offers both knowledge...

Computer games represent an ideal research domain for the next generation of personalized digital applications. This paper presents a player-centered framework AI game personalization, complementary to commonly used system-centered approaches. Built on Structure Actions theory, maps out current landscape personalization and identifies eight open problems that need further investigation. These require deep collaboration between technological advancement player experience design.

Social comparison-based features are widely used in social computing apps. However, most existing apps not grounded comparison theories and do consider individual differences preferences reactions. This paper is among the first to automatically personalize targets. In context of an m-health app for physical activity, we use artificial intelligence (AI) techniques multi-armed bandits. Results from our user study (n=53) indicate that there some evidence motivation can be increased using...

Educational technology is shifting toward facilitating personalized learning. Such personalization, however, requires a detailed understanding of students’ problem-solving processes. Sequence analysis (SA) promising approach to gaining granular insights into student problem solving; existing techniques are difficult interpret because they offer little room for human input in the process. Ultimately, learning context, stakeholder makes decisions, so should be able drive In this paper, we...

Background/Purpose: Adolescents from urban, socioeconomically disadvantaged communities of color encounter high rates adverse childhood experiences. To address the resulting multidimensional problems, we developed an innovative approach, Experiential Participatory and Interactive Knowledge Elicitation (EPIKE), using remote experiential needs elicitation methods to generate design content requirements for a mobile health (mHealth) psychoeducational intervention. Methods: At community-based...

In this paper, we present the iterative design of StepQuest, a Fitbit-based social motion-based game for health (MGH) to sustain physical activity (PA) and support extended play. We conducted two 6-week user studies (n=24) evaluate effectiveness promote PA an period time as well role existing relationship. Our findings indicate that pre-existing positive relationship (e.g., friendship) has impact on players' levels when they play MGH, compared strangers, effect was amplified more gameplay...

This paper focuses on building personalized player models solely from behavior in the context of adaptive games. We present two main contributions: The first is a novel approach to modeling based multi-armed bandits (MABs). addresses, at same time and principled way, both problem collecting data model characteristics interest for current adapting interactive experience this model. Second, we an evaluating fine-tuning these algorithms prior generating user study. important problem, because...

Background Innovative approaches are needed to understand barriers and facilitators of physical activity among insufficiently active adults. Although social comparison processes (ie, self-evaluations relative others) often used motivate in digital environments, user preferences responses information poorly understood. Objective We an iterative approach better users’ selection targets, how they interacted with their selected responded these targets. Methods Across 3 studies, different samples...

This paper focuses on tracing player knowledge in educational games. Specifically, given a set of concepts or skills required to master game, the goal is estimate likelihood with which current has mastery each those skills. The main contribution an approach that integrates machine learning and domain rules find when applied certain skill either succeeded failed. then as input standard module (such from Intelligent Tutoring Systems) perform tracing. We evaluate our context game called...

Reflection is a critical aspect of the learning process. However, educational games tend to focus on supporting concepts rather than reflection. While reflection occurs in games, game design and research community can benefit from more knowledge how facilitate player through design. In this paper, we examine programming analyze currently supported. We find that current approaches prioritize accuracy over individual process often only support post-gameplay. Our analysis identifies common...

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 story-based map generation focusing on the tight relationship between virtual where those will unfold. Our long term goal is develop content techniques that can produce maps supporting multiple stories. We present an takes, as input, specification of story space collection plot points. Causal relations these points spatial...

Modern computing is increasingly handled in a parallel fashion, however, little known about how individuals learn programming. This paper focuses on the design of an educational game called Parallel, designed for both teaching programming education CS undergraduate curricula, as well gathering insights into students learn, and solve problems. Specifically, we focus key challenge choosing appropriate metaphors order to facilitate transference between game. In this paper, describe our...