Papers by Peter Brusilovsky
The emergence of generative large language models (LLMs) raises the question: what will be its im... more The emergence of generative large language models (LLMs) raises the question: what will be its impact on crowdsourcing? Traditionally, crowdsourcing has been used for acquiring solutions to a wide variety of human-intelligence tasks, including ones involving text generation, modification or evaluation. For some of these tasks, models like ChatGPT can potentially substitute human workers. In this study, we investigate whether this is the case for the task of paraphrase generation for intent classification. We apply data collection methodology of an existing crowdsourcing study (similar scale, prompts and seed data) using ChatGPT and Falcon-40B. We show that ChatGPT-created paraphrases are more diverse and lead to at least as robust models.
Springer eBooks, Jul 28, 2008
Springer eBooks, May 14, 2007
Adaptive navigation support is a specific group of technologies that support user navigation in h... more Adaptive navigation support is a specific group of technologies that support user navigation in hyperspace, by adapting to the goals, preferences and knowledge of the individual user. These technologies, originally developed in the field of adaptive hypermedia, are becoming increasingly important in several adaptive Web applications, ranging from Web-based adaptive hypermedia to adaptive virtual reality. This chapter provides a brief introduction to adaptive navigation support, reviews major adaptive navigation support technologies and mechanisms, and illustrates these with a range of examples.
User Modeling and User-adapted Interaction, Jul 1, 1996
In this talk I will introduce the emerging technology of Open Social Student Modeling (OSSM) and ... more In this talk I will introduce the emerging technology of Open Social Student Modeling (OSSM) and review several projects performed in our research lab to investigate the potential of OSSM. OSSM is a recent extension of Open Student Modeling (OSM), a popular technology in the area of personalized learning systems. While in traditional personalized systems, student models were hidden "under the hood" and used to personalize the educational process; open student modeling introduced the ability to view and modify the state of students' own knowledge to support reflection, self-organized learning, and system transparency. Open Social Student Modeling takes this idea one step further by allowing students to explore each other's models or an aggregated model of the class. The idea to make OSM social was originally suggested and explored by Bull [1; 2]. Over the last few years, our team explored several approaches to present OSSM in a highly visual form and evaluated these approaches in a sequence of classroom and lab studies. I will present a summary of this work introducing such systems as QuizMap [3], Progressor [4], and Mastery Grids [5] and reviewing most interesting research evidence collected by the studies.
The New Review of Hypermedia and Multimedia, Apr 1, 2009
Already in the earliest ideas leading up to hypermedia, starting with Vannevar Bush's Memex (Bush... more Already in the earliest ideas leading up to hypermedia, starting with Vannevar Bush's Memex (Bush 1945), the personal view of and access to and through information was a key aspect of hypertext and hypermedia research. In traditional hypertext, personalization was user-driven: the existence of links enabled each user to follow his/her own path through the hyperspace. However, since about 1990, the topic of system-driven personalization has started to receive more attention. Instead of just the user selecting a personal path through hyperspace personalized, dynamic and/or adaptive hyperdocuments change the hyperspace itself as the user is traversing (or otherwise using) it. The mid-1990s mark the start of serious research into hypertext, hypermedia, and later Web personalization and adaptation and the emergence of an active adaptive hypermedia community with workshops on adaptive hypermedia at the User Modeling, the ACM Hypertext and the World Wide Web conferences. This has culminated in the start of a separate biannual conference series on Adaptive Hypermedia, the first one in Trento, Italy, in August 2000. Adaptive Hypermedia conferences alternated with the User Modeling conferences, showing, year by year, an increasing overlap in interest and attendance as adaptation heavily relies on user modeling and user modeling is almost always done in order to achieve some form of adaptation or personalization. The year 2008 marked the end of the separate conference series, and the merger of the Adaptive Hypermedia and User Modeling events into a new annual conference on User Modeling, Adaptation and Personalization. Adaptive Hypermedia has come of age and become an important component of the global research on adaptation and personalization. This special issue, third in the line of NRHM special issues devoted to Adaptive Hypermedia (after NRHM 4, 1998 and NRHM 10(1), 2004) attempts to present a snapshot of Adaptive Hypermedia research in this important point of time. Building on recent Adaptive Hypermedia conferences and ACM Hypertext conferences, this special issue brings together an interesting mix of recent developments. Even though unplanned at the time of soliciting papers, we ended up with an outstanding mix of AH-related topics, describing technical issues as well as the usability of adaptive applications. The first paper in this special issue brings together the past and future. Appropriately titled AH 12 years later: a comprehensive survey of adaptive hypermedia methods and techniques (Knutov, De Bra, Pechenizkiy), it reflects on the AH research since the seminal paper (Brusilovsky 1996) ''Methods and techniques of adaptive hypermedia'' that has become a standard reference in
British Journal of Educational Technology, Aug 1, 2003
Adaptive navigation support (ANS) is a new direction of research within the area of adaptive inte... more Adaptive navigation support (ANS) is a new direction of research within the area of adaptive interfaces. The goal of ANS techniques is to help users find an appropriate path in the learning and information space by adapting link presentation to the goals, knowledge, and other characteristics of an individual user. This paper is devoted to evaluation of adaptive navigation support in educational context. We present an educational hypermedia system ISIS-Tutor that applies several ANS technologies-adaptive annotation, adaptive hiding, and direct guidance-and describe a study, which evaluates the first two technologies. The results show that adaptive navigation support is helpful and can reduce user navigation efforts.
Knowledge Discovery and Data Mining, 2019
Identifying experts is an important challenge in many contexts. In this paper, we present a metho... more Identifying experts is an important challenge in many contexts. In this paper, we present a method to build a knowledge graph by integrating data from Google Scholar and Wikipedia to help students find a research advisor or thesis committee member. This knowledge graph is used to power the exploratory search interface to recommend similar keywords and relevant scholars to the students with a limited level of knowledge and familiarity with the subject of research.
In this talk, I will review our research attempts to implement different kinds of personalization... more In this talk, I will review our research attempts to implement different kinds of personalization in the context of relevance-based visualization. The goal of this research stream is to make relevance-based visualization adaptive to user long-term goals, interests, or prospects rather just responsive to short term immediate needs such as query terms. I will present four personalized relevance-based visualization systems: Adaptive VIBE, TalkExplorer, SetFusion, and IntersectionExplorer, For each system, I will present its idea, some evaluation results, and lessons learned.
Zenodo (CERN European Organization for Nuclear Research), Mar 22, 2023
Self-efficacy, or the belief in one's ability to accomplish a task or achieve a goal, can signifi... more Self-efficacy, or the belief in one's ability to accomplish a task or achieve a goal, can significantly influence the effectiveness of various instructional methods to induce learning gains. The importance of self-efficacy is particularly pronounced in complex subjects like Computer Science, where students with high self-efficacy are more likely to feel confident in their ability to learn and succeed. Conversely, those with low self-efficacy may become discouraged and consider abandoning the field. The work presented here examines the relationship between self-efficacy and students learning computer programming concepts. For this purpose, we conducted a randomized control trial experiment with university-level students who were randomly assigned into two groups: a control group where participants read Java programs accompanied by explanatory texts (a passive strategy) and an experimental group where participants self-explain while interacting through dialogue with an intelligent tutoring system (an interactive strategy). We report here the findings of this experiment with a focus on self-efficacy, its relation to students' learning gains (to evaluate the effectiveness, we measure pre/post-test), and other important factors such as prior knowledge or experimental condition/instructional strategies as well as interaction effects.
IEEE Transactions on Learning Technologies, Apr 1, 2014
W ELCOME to the second 2014 issue of the IEEE Transactions on Learning Technologies. This issue p... more W ELCOME to the second 2014 issue of the IEEE Transactions on Learning Technologies. This issue presents a collection of nine papers focused, as in the majority of our issues, on a broad set of learning technologies. The issue starts with two papers that address the design of augmented reality systems for learning. In "Computer-Assisted Culture Learning in an Online Augmented Reality Environment Based on Free-Hand Gesture Interaction," Mau-Tsuen Yang and Wan-Che Liao apply immersive augmented reality to cultural language learning. They have devised a system named VECAR, where a learner wearing a head-mounted display can interact with a map, or an image of a heritage site or cultural object, through hand and finger gestures. They evaluated the system against a control group using a Google map, for learning effectiveness, communication between teacher and students, and usability and learnability. The VECAR condition produced higher learning gains and enhanced communication with the teacher. Despite some technical difficulties, learners reported an engaging experience.
His research interests are centered around adaptive Web-based systems, user modeling, intelligent... more His research interests are centered around adaptive Web-based systems, user modeling, intelligent tutoring systems, and adaptive hypermedia. For the last five years he has been involved in several projects related to developing adaptive systems on the Web and integration of different Web-based adaptive systems. He is an author of many papers and an editor of several books related to the topic of the workshop.
Zenodo (CERN European Organization for Nuclear Research), Mar 15, 2023
Zenodo (CERN European Organization for Nuclear Research), Mar 22, 2023
Proceedings of the 25th International Conference on Intelligent User Interfaces Companion
We demonstrate an intelligent, personalized, multifaceted visualization of people recommendation ... more We demonstrate an intelligent, personalized, multifaceted visualization of people recommendation using a personalized 2D entities graph and a word cloud for exploration by the user. This visualization aims to show non-trivial connections, e.g., those that the user may had forgotten about, but they are interesting and relevant. Since entities we are linked to are part of our lives (and profile), they help to understand who we are and what are we interested in. We adapt the typed entity-relation graph (profile) concept as introduced by [1] and based on this presentation we visualize the entity profile. In this demonstration, the users, as case study are the participants of IUI'20, will be able to explore their own personalized entities graph based on entities and relations that the system harvest about them (after getting their approval), from the web for finding interesting connections that they may meet in the context of this conference.
Zenodo (CERN European Organization for Nuclear Research), Jul 18, 2022
Computing is an increasingly fundamental skill for students across disciplines. It enables them t... more Computing is an increasingly fundamental skill for students across disciplines. It enables them to solve complex, real and challenging problems and make a positive impact in the world. Yet, the field of computing education is still facing a range of problems from high failure and attrition rates, to challenges in training and recruiting teachers, to the under-representation of women and students of color. Do not delete, move, or resize this block. If the paper is accepted, this block will need to be filled in with reference information.
2020 IEEE 8th International Conference on Serious Games and Applications for Health (SeGAH), 2020
One of the shortcomings of many modern serious games and medical simulations lies in their inabil... more One of the shortcomings of many modern serious games and medical simulations lies in their inability to model even some modicum of unpredictability of real life situations. Interactions with a standardized patient may teach healthcare professional students how to diagnose a clinical condition, better manage a patient, or help them improve their bedside manners, but such simulated interactions will not prepare the learners to deal with unpredictability of clinical situations, interruption, and task switching. Distractions occur from colleagues, clinical decision support alerts, pagers, smartphones, or audible alarms. All these interruptions can potentially alter the course of patient care and the outcome of a patient's treatment. A simulated virtual patient (VP) may teach critical thinking skills, but once a student has successfully diagnosed a VP, the simulation stops providing educational value. In this paper we propose a generalizable method for integrating agent-based models into serious games and simulations. In the proposed paradigm, a human player (learner) takes on the role of a single agent in the model (e.g, a healthcare professional), while the output of the model controls the environment, the rules of agent interactions, and all the other agents that the human player interacts with (non-player characters). Moreover, we will present two use cases demonstrating that the use of agent-based models as behavior controllers for non-player characters introduces a degree of unpredictability in a virtual patient simulation and in a serious game designed to teach middle and high-school students about the spread of infectious diseases.
Adjunct Proceedings of the 29th ACM Conference on User Modeling, Adaptation and Personalization, 2021
An emerging challenge in course recommender systems is explaining to students why they have been ... more An emerging challenge in course recommender systems is explaining to students why they have been recommended particular courses. In the context of a university, it can be valuable for a recommender system to introduce students to courses they may have not otherwise taken but which are still relevant to them. However, there is a tension between these goals as students have less ability to judge the relevance of courses, the less familiar they are with them. In this paper, we explore ways of familiarizing students with recommendations with three types of explanations designed with varying levels of personalization. We conduct a 67 student randomized controlled experiment using two course recommendation engines, content and context-based, and augment recommendations with three different types of explanation. Students rated each course recommended in terms of novelty, unexpectedness, and successfulness (i.e., intent to enroll). We find several statistically significant results, including an increase in serendipity (i.e., unexpectedness + successfulness) when explaining new course recommendations using keywords from courses a student has previously taken.
Uploads
Papers by Peter Brusilovsky