CityFlocks: Designing Social Navigation for Urban Mobile Information Systemskavasmlikon
CityFlocks is a mobile system enabling visitors and new residents in a city to tap into the knowledge and experiences of local residents, so as to gather information about their new environment. Its design specifically aims to lower existing barriers of access and facilitate social navigation in urban places. This paper presents a design case study of a mobile system prototype that offers an easy way for information seeking new residents or visitors to access tacit knowledge from local people about their new community. In various user tests we evaluate two general user interaction alternatives – direct and indirect social navigation – and analyse under what conditions which interaction method works better for people using a mobile device to socially navigate urban environments. The outcomes are relevant for the user interaction design of future mobile information systems that leverage off of a social navigation approach.
Empirical studies of adaptive annotation in the educational context have demonstrated that it can help students to acquire knowledge faster, improve learning outcomes, reduce navigational overhead, and encourage non-sequential navigation. Over the last 8 years we have explored a lesser known effect of adaptive annotation – its ability to significantly increase student engagement in working with non-mandatory educational content. In the presence of adaptive link annotation, students tend to access significantly more learning content; they stay with it longer, return to it more often and explore a wider variety of learning resources. This talk will present an overview of our exploration of the addictive links effect in many course-long studies, which we ran in several domains (C, SQL and Java programming), for several types of learning content (quizzes, problems, interactive examples). The first part of the talk will review our exploration of a more traditional knowledge-based personalization approach and the second part will focus on more recent studies of social navigation and open social student modeling
presentation at ICALT2012 of the paper "Falakmasir M., Hsiao I., Mazzola L., Grant N., Brusilovsky P. (2012). The Impact of Social
Performance Visualization on Students. Proceedings of the 12th IEEE International
Conference on Advanced Learning Technologies and Technology-enhanced Learning, Rome,
Italy, July 4-6, 2012, pgg. 565-569"
Program code examples (known also as worked examples) play a crucial role in learning how to program. Instructors use examples extensively to demonstrate the semantics of the programming language being taught and to highlight the fundamental coding patterns. Programming textbooks allocate considerable space to present and explain code examples. To make the process of studying code examples more interactive, CS education researchers developed a range of tools to engage students in the study of code examples. These tools include codecasts (codemotion,codecast,elicasts), interactive example explorers (WebEx, PCEX), and tutoring systems (DeepTutor). An important component in all types of worked examples is code explanations associated with specific code lines or code chunks of an example. The explanations connect examples with general programming knowledge explaining the role and function of code fragments or their behavior. In textbooks, these explanations are usually presented as comments in the code or as explanations on the margins. The example explorer tools allow students to examine these explanations interactively. Tutoring systems, which engage students in explaining the code, use these model explanations to check student responses and provide scaffolding. In all these cases, to make a worked example re-usable beyond its presentation in a lecture, the explanations have to be authored by instructors or domain experts i.e., produced and integrated into a specific system. As the experience of the last 10 years demonstrated, these explanations are hard to obtain. Those already collected are usually “locked” in a specific example-focused system and can’t be reused. The purpose of this working group is to support broader re-used of worked examples augmented with explanations. Our current plan is to develop а standard approach to represent explained examples. This approach will enable an example created for any of the existing systems to be explored in a standard format and imported into any other example-focused system. We plan to follow a successful experience of the PEML working group focused on re-using programming exercises.
SANN: Programming Code Representation Using Attention Neural Network with Opt...Peter Brusilovsky
Slides of CIKM 2023 paper by Muntasir Hoq, Sushanth Reddy Chilla, Melika Ahmadi Ranjbar, Peter Brusilovsky and Bita Akram
https://dl.acm.org/doi/10.1145/3583780.3615047
CityFlocks: Designing Social Navigation for Urban Mobile Information Systemskavasmlikon
CityFlocks is a mobile system enabling visitors and new residents in a city to tap into the knowledge and experiences of local residents, so as to gather information about their new environment. Its design specifically aims to lower existing barriers of access and facilitate social navigation in urban places. This paper presents a design case study of a mobile system prototype that offers an easy way for information seeking new residents or visitors to access tacit knowledge from local people about their new community. In various user tests we evaluate two general user interaction alternatives – direct and indirect social navigation – and analyse under what conditions which interaction method works better for people using a mobile device to socially navigate urban environments. The outcomes are relevant for the user interaction design of future mobile information systems that leverage off of a social navigation approach.
Empirical studies of adaptive annotation in the educational context have demonstrated that it can help students to acquire knowledge faster, improve learning outcomes, reduce navigational overhead, and encourage non-sequential navigation. Over the last 8 years we have explored a lesser known effect of adaptive annotation – its ability to significantly increase student engagement in working with non-mandatory educational content. In the presence of adaptive link annotation, students tend to access significantly more learning content; they stay with it longer, return to it more often and explore a wider variety of learning resources. This talk will present an overview of our exploration of the addictive links effect in many course-long studies, which we ran in several domains (C, SQL and Java programming), for several types of learning content (quizzes, problems, interactive examples). The first part of the talk will review our exploration of a more traditional knowledge-based personalization approach and the second part will focus on more recent studies of social navigation and open social student modeling
presentation at ICALT2012 of the paper "Falakmasir M., Hsiao I., Mazzola L., Grant N., Brusilovsky P. (2012). The Impact of Social
Performance Visualization on Students. Proceedings of the 12th IEEE International
Conference on Advanced Learning Technologies and Technology-enhanced Learning, Rome,
Italy, July 4-6, 2012, pgg. 565-569"
Program code examples (known also as worked examples) play a crucial role in learning how to program. Instructors use examples extensively to demonstrate the semantics of the programming language being taught and to highlight the fundamental coding patterns. Programming textbooks allocate considerable space to present and explain code examples. To make the process of studying code examples more interactive, CS education researchers developed a range of tools to engage students in the study of code examples. These tools include codecasts (codemotion,codecast,elicasts), interactive example explorers (WebEx, PCEX), and tutoring systems (DeepTutor). An important component in all types of worked examples is code explanations associated with specific code lines or code chunks of an example. The explanations connect examples with general programming knowledge explaining the role and function of code fragments or their behavior. In textbooks, these explanations are usually presented as comments in the code or as explanations on the margins. The example explorer tools allow students to examine these explanations interactively. Tutoring systems, which engage students in explaining the code, use these model explanations to check student responses and provide scaffolding. In all these cases, to make a worked example re-usable beyond its presentation in a lecture, the explanations have to be authored by instructors or domain experts i.e., produced and integrated into a specific system. As the experience of the last 10 years demonstrated, these explanations are hard to obtain. Those already collected are usually “locked” in a specific example-focused system and can’t be reused. The purpose of this working group is to support broader re-used of worked examples augmented with explanations. Our current plan is to develop а standard approach to represent explained examples. This approach will enable an example created for any of the existing systems to be explored in a standard format and imported into any other example-focused system. We plan to follow a successful experience of the PEML working group focused on re-using programming exercises.
SANN: Programming Code Representation Using Attention Neural Network with Opt...Peter Brusilovsky
Slides of CIKM 2023 paper by Muntasir Hoq, Sushanth Reddy Chilla, Melika Ahmadi Ranjbar, Peter Brusilovsky and Bita Akram
https://dl.acm.org/doi/10.1145/3583780.3615047
Personalized Learning: Expanding the Social Impact of AIPeter Brusilovsky
Slide of my keynote talk at SIAIA '23 workshop held at AAAI 2023:
The use of AI in Education could be traced to the early days of AI. While the publicity associated with the most recent wave of AI applications rarely mentions education, it is through the improvement in education AI could achieve an impressive social impact. In particular, the AI ability to personalize the learning process could make a large difference in a context where learners' knowledge could be radically different from learner to learner. Modern computer and internet technologies can now bring the power of learning in the forms of MOOCs, online textbooks, and zoom courses truly worldwide. Yet, without personalization, the potential of these technologies is not fully leveraged. In this talk, I will review several generations of research on personalized learning and discuss tools, technologies, and infrastructures for personalized learning that we are currently exploring.
Action Sequence Mining and Behavior Pattern Analysis for User ModelingPeter Brusilovsky
Slides of my talk at 2022 Workshop on Temporal Aspects of User Modeling
Tracing learner interaction with educational content has recently emerged as a centerpiece of learning analytics. Augmented by various data mining technologies, learner data has been used to predict learner success and failure, prevent dropouts, and inform university officials about student progress. While the majority of existing learning analytics approaches ignore the time aspect in the learning data, recent research indicated that not just what the learners do, but how and in which order they do it is critical to understand differences between learners, model their behavior, and predict their performance. In my talk, I will focus on the application of action sequence mining as a tool to extract temporal patterns of learning behavior and recognize cohorts of learners with divergent behavior. I will review three case studies of using sequence mining with learner data, present the obtained results, and discuss their importance for user modeling and personalization.
Tutorial at UMAP 2022:
In recent years, the use of Artificial Intelligence (AI) technologies expanded to many areas where they directly affect the lives of many
people. AI-based approaches advise human decision-makers who should be released on bail, whether it is a good time to discharge a
patient from a hospital and whether a specific student is at risk to fail a course. Such an extensive use in AI in decision making came with
a range of protentional problems that have been extensively studied over the last few years. Recognition of these problems motivated a
rapid rise of research on “human-centered AI”, which attempted to address and minimize the negative effects of using AI technologies.
Among the ideas of human-centered AI is user control - engaging users in affecting AI decision making to prevent possible errors and
biases. In my talk, I will focus on the application of user control in one popular area of AI application, adaptive information access.
Adaptive information access systems such as personalized search and recommender systems attempt to model their users to help them in
finding the most relevant information. Yet, user modeling and personalization mechanisms might not always work as expected resulting
in errors, biases, and suboptimal behavior. Combining the decision power or AI with the ability of the user to guide and control it brings
together the strong sides of artificial and human intelligence and could lead to better results. This tutorial will provide a systematic review
of approaches focused on adding various kinds of user control to adaptive information access systems and discuss lessons learned,
prospects, and challenges of this direction of research.
Human-Centered AI in AI-ED - Keynote at AAAI 2022 AI for Education workshopPeter Brusilovsky
Abstract: In recent years, the use of Artificial Intelligence (AI) technologies expanded to many areas directly affecting the lives of millions. AI-based approaches advise human decision-makers who should be released on bail, whether it is a good time to discharge a patient from a hospital and whether a specific student is at risk to fail a course. Such extensive use in AI in decision making came with a range of protentional problems that have been extensively studied over the last few years. Recognition of these problems motivated a rapid rise of research on “human-centered AI”, which attempted to address and minimize the negative effects of using AI technologies. The majority of work on human-centered AI focus on various types of Human-AI collaboration through such technologies as transparency, explainability, and user control. In my talk, I will review how the ideas of Human-AI collaboration, transparency, explainability, and user control have been used in educational applications of AI in the past and will discuss now new ideas in this research area developed outside of AI-Ed could be creatively applied in educational context.
User Control in AIED (Artificial Intelligence in Education)Peter Brusilovsky
Slides of my intro to "Meet the Expert" session at AIED 2021. This is a subset of slides of a longer presentation on user control in AI extended with many specific examples from AIED area.
The Return of Intelligent Textbooks - ITS 2021 keynote talkPeter Brusilovsky
Early research on hypermedia learning and Web-based education featured a strong stream of work on intelligent and adaptive textbooks, which combined the knowledge modeling ideas from the field of intelligent tutoring with rich linking offered by the hypermedia and the Web. However, over the next ten years from 2005 to 2015 this area was relatively quiet as the focus of research in e-learning has shifted to other topics and other creative ideas to leverage the power of Internet. A recent gradual shift of the whole publication industry from printed books to electronic books followed by a rapid growth or the volume of online books re-ignited interests to “more intelligent” textbooks. The research on the new generation of intelligent textbooks engaged a larger set of technologies and engaged scholars from a broader range of areas including machine learning, natural language understanding, social computing, etc. In my talk I will review the past and present of research on intelligent textbooks from its origins to the diverse modern work providing examples of most interesting technologies and research results.
Data-Driven Education 2020: Using Big Educational Data to Improve Teaching an...Peter Brusilovsky
Modern educational settings from regular classrooms to MOOCs produce a a rapidly increasing volume of data that captures individual learning progress of millions of students at different level of granularity. This presence of this data opens a unique opportunity to re-engineer traditional education and build and develop a range of efficient data-driven approaches to support teaching and learning. In my talk, I will present several ways to use big educational data explored in our lab. The focus will be on open social learning modeling and identifying individual differences through sequential pattern mining, but several other approaches will be mentioned. Open social learning modeling and sequential pattern mining provides two considerably different examples on using educational data. One offers an immediate use of class interaction history to develop more engaging content access while another shows how big data could be used to uncover important individual differences that could be used to optimize the process for individual leaners.
Two Brains are Better than One: User Control in Adaptive Information AccessPeter Brusilovsky
In recent years, the use of Artificial Intelligence (AI) technologies expanded to many areas where they directly affect the lives of many people. AI-based approaches advise human decision-makers who should be released on bail, whether it is a good time to discharge a patient from a hospital and whether a specific student is at risk to fail a course. Such an extensive use in AI in decision making came with a range of protentional problems that have been extensively studied over the last few years. Recognition of these problems motivated a rapid rise of research on “human-centered AI”, which attempted to address and minimize the negative effects of using AI technologies. Among the ideas of human-centered AI is user control - engaging users in affecting AI decision making to prevent possible errors and biases. In my talk, I will focus on the application of user control in one popular area of AI application, adaptive information access. Adaptive information access systems such as personalized search and recommender systems attempt to model their users to help them in finding the most relevant information. Yet, user modeling and personalization mechanisms might not always work as expected resulting in errors, biases, and suboptimal behavior. Combining the decision power or AI with the ability of the user to guide and control it brings together the strong sides of artificial and human intelligence and could lead to better results. In my talk, I review several projects focused on user control in adaptive information access systems and discuss the benefits and challenges of this approach.
Personalized Online Practice Systems for Learning ProgrammingPeter Brusilovsky
Computer programming is quickly transitioning from being just a key competency in computer and information science majors to being a desired skill for students in a wide range of fields. Yet, it is also one of the most challenging subjects to learn. While learning by doing is a critical component in mastering programming skills, neither the traditional educational process nor standard learning support tools provide sufficient opportunities for programming practice. In this talk, I will present our research on personalized programming practice systems for Java, Python, and SQL, which attempt to bridge this known gap in learning programming. A programming practice system engages students in practicing programming skills beyond a relatively small number of graded assignments and exams. To support learning by doing, an online practice system offers a range of interactive “smart content” such as program animations, worked examples, and various kinds of programming problems with an automatic assessment. The main challenges for online practice systems are to motivate students to practice and to guide them to the most appropriate smart content given their course goals and knowledge levels. In this talk, I will review a range of AI technologies, such as student modeling, navigation support, social comparison, and content recommendation, which support efficient programming practice. I will also discuss how personalized practice system could support COVID-19-influenced switch to online learning while maintaining an extensive level of feedback expected from an efficient learning process.
UMAP 2019 talk Evaluating Visual Explanations for Similarity-Based Recommenda...Peter Brusilovsky
Tsai, Chun-Hua, and Peter Brusilovsky. 2019. "Evaluating Visual Explanations for Similarity-Based Recommendations: User Perception and Performance." In the 27th ACM Conference on User Modeling, Adaptation and Personalization, UMAP 2019, 22-30. Larnaca, Cyprus: ACM.
Course-Adaptive Content Recommender for Course AuthoringPeter Brusilovsky
Developing online courses is a complex and time-consuming
process that involves organizing a course into a sequence of topics and
allocating the appropriate learning content within each topic. This task
is especially difficult in complex domains like programming, due to the
incremental nature of programming knowledge, where new topics extensively
build upon domain concepts that were introduced in earlier lessons.
In this paper, we propose a course-adaptive content-based recommender
system that assists course authors and instructors in selecting the most
relevant learning material for each course topic. The recommender system
adapts to the deep prerequisite structure of the course as envisioned
by a specific instructor, while unobtrusively deducing that structure from
problem-solving examples that the instructor uses to present course concepts.
We assessed the quality of recommendations and examined several
aspects of the recommendation process by using three datasets collected
from two different courses.While the presented recommender system was
built for the domain of introductory programming, our course-adaptive
recommendation approach could be used in a variety of other domains.
Data-Driven Education: Using Big Educational Data to Improve Teaching and Learning. Keynote slides for 15th International Conference on Web-Based Learning, ICWL 2016, Rome, Italy, October 26–29.
From Expert-Driven to Data-Driven Adaptive LearningPeter Brusilovsky
Keynote slides for the Workshop on Advancing Education with Data at the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Halifax, NS, Canada, Aug 14, 2017
Stereotype Modeling for Problem-Solving Performance Predictions in MOOCs and ...Peter Brusilovsky
Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization, UMAP2017, pp 76-84
Stereotypes are frequently used in real life to classify students according to their performance in class. In literature, we can find many references to weaker students, fast learners, struggling students, etc. Given the lack of detailed data about students, these or other kinds of stereotypes could be potentially used for user modeling and personalization in the educational context. Recent research in MOOC context demonstrated that data-driven learner stereotypes could work well for detecting and preventing student dropouts. In this paper, we are exploring the application of stereotype-based modeling to a more challenging task -- predicting student problem-solving and learning in two programming courses and two MOOCs. We explore traditional stereotypes based on readily available factors like gender or education level as well as some advanced data-driven approaches to group students based on their problem-solving behavior. Each of the approaches to form student stereotype cohorts is validated by comparing models of student learning: do students in different groups learn differently? In the search for the stereotypes that could be used for adaptation, the paper examines ten approaches. We compare the performance of these approaches and draw conclusions for future research.
Personalized Learning: Expanding the Social Impact of AIPeter Brusilovsky
Slide of my keynote talk at SIAIA '23 workshop held at AAAI 2023:
The use of AI in Education could be traced to the early days of AI. While the publicity associated with the most recent wave of AI applications rarely mentions education, it is through the improvement in education AI could achieve an impressive social impact. In particular, the AI ability to personalize the learning process could make a large difference in a context where learners' knowledge could be radically different from learner to learner. Modern computer and internet technologies can now bring the power of learning in the forms of MOOCs, online textbooks, and zoom courses truly worldwide. Yet, without personalization, the potential of these technologies is not fully leveraged. In this talk, I will review several generations of research on personalized learning and discuss tools, technologies, and infrastructures for personalized learning that we are currently exploring.
Action Sequence Mining and Behavior Pattern Analysis for User ModelingPeter Brusilovsky
Slides of my talk at 2022 Workshop on Temporal Aspects of User Modeling
Tracing learner interaction with educational content has recently emerged as a centerpiece of learning analytics. Augmented by various data mining technologies, learner data has been used to predict learner success and failure, prevent dropouts, and inform university officials about student progress. While the majority of existing learning analytics approaches ignore the time aspect in the learning data, recent research indicated that not just what the learners do, but how and in which order they do it is critical to understand differences between learners, model their behavior, and predict their performance. In my talk, I will focus on the application of action sequence mining as a tool to extract temporal patterns of learning behavior and recognize cohorts of learners with divergent behavior. I will review three case studies of using sequence mining with learner data, present the obtained results, and discuss their importance for user modeling and personalization.
Tutorial at UMAP 2022:
In recent years, the use of Artificial Intelligence (AI) technologies expanded to many areas where they directly affect the lives of many
people. AI-based approaches advise human decision-makers who should be released on bail, whether it is a good time to discharge a
patient from a hospital and whether a specific student is at risk to fail a course. Such an extensive use in AI in decision making came with
a range of protentional problems that have been extensively studied over the last few years. Recognition of these problems motivated a
rapid rise of research on “human-centered AI”, which attempted to address and minimize the negative effects of using AI technologies.
Among the ideas of human-centered AI is user control - engaging users in affecting AI decision making to prevent possible errors and
biases. In my talk, I will focus on the application of user control in one popular area of AI application, adaptive information access.
Adaptive information access systems such as personalized search and recommender systems attempt to model their users to help them in
finding the most relevant information. Yet, user modeling and personalization mechanisms might not always work as expected resulting
in errors, biases, and suboptimal behavior. Combining the decision power or AI with the ability of the user to guide and control it brings
together the strong sides of artificial and human intelligence and could lead to better results. This tutorial will provide a systematic review
of approaches focused on adding various kinds of user control to adaptive information access systems and discuss lessons learned,
prospects, and challenges of this direction of research.
Human-Centered AI in AI-ED - Keynote at AAAI 2022 AI for Education workshopPeter Brusilovsky
Abstract: In recent years, the use of Artificial Intelligence (AI) technologies expanded to many areas directly affecting the lives of millions. AI-based approaches advise human decision-makers who should be released on bail, whether it is a good time to discharge a patient from a hospital and whether a specific student is at risk to fail a course. Such extensive use in AI in decision making came with a range of protentional problems that have been extensively studied over the last few years. Recognition of these problems motivated a rapid rise of research on “human-centered AI”, which attempted to address and minimize the negative effects of using AI technologies. The majority of work on human-centered AI focus on various types of Human-AI collaboration through such technologies as transparency, explainability, and user control. In my talk, I will review how the ideas of Human-AI collaboration, transparency, explainability, and user control have been used in educational applications of AI in the past and will discuss now new ideas in this research area developed outside of AI-Ed could be creatively applied in educational context.
User Control in AIED (Artificial Intelligence in Education)Peter Brusilovsky
Slides of my intro to "Meet the Expert" session at AIED 2021. This is a subset of slides of a longer presentation on user control in AI extended with many specific examples from AIED area.
The Return of Intelligent Textbooks - ITS 2021 keynote talkPeter Brusilovsky
Early research on hypermedia learning and Web-based education featured a strong stream of work on intelligent and adaptive textbooks, which combined the knowledge modeling ideas from the field of intelligent tutoring with rich linking offered by the hypermedia and the Web. However, over the next ten years from 2005 to 2015 this area was relatively quiet as the focus of research in e-learning has shifted to other topics and other creative ideas to leverage the power of Internet. A recent gradual shift of the whole publication industry from printed books to electronic books followed by a rapid growth or the volume of online books re-ignited interests to “more intelligent” textbooks. The research on the new generation of intelligent textbooks engaged a larger set of technologies and engaged scholars from a broader range of areas including machine learning, natural language understanding, social computing, etc. In my talk I will review the past and present of research on intelligent textbooks from its origins to the diverse modern work providing examples of most interesting technologies and research results.
Data-Driven Education 2020: Using Big Educational Data to Improve Teaching an...Peter Brusilovsky
Modern educational settings from regular classrooms to MOOCs produce a a rapidly increasing volume of data that captures individual learning progress of millions of students at different level of granularity. This presence of this data opens a unique opportunity to re-engineer traditional education and build and develop a range of efficient data-driven approaches to support teaching and learning. In my talk, I will present several ways to use big educational data explored in our lab. The focus will be on open social learning modeling and identifying individual differences through sequential pattern mining, but several other approaches will be mentioned. Open social learning modeling and sequential pattern mining provides two considerably different examples on using educational data. One offers an immediate use of class interaction history to develop more engaging content access while another shows how big data could be used to uncover important individual differences that could be used to optimize the process for individual leaners.
Two Brains are Better than One: User Control in Adaptive Information AccessPeter Brusilovsky
In recent years, the use of Artificial Intelligence (AI) technologies expanded to many areas where they directly affect the lives of many people. AI-based approaches advise human decision-makers who should be released on bail, whether it is a good time to discharge a patient from a hospital and whether a specific student is at risk to fail a course. Such an extensive use in AI in decision making came with a range of protentional problems that have been extensively studied over the last few years. Recognition of these problems motivated a rapid rise of research on “human-centered AI”, which attempted to address and minimize the negative effects of using AI technologies. Among the ideas of human-centered AI is user control - engaging users in affecting AI decision making to prevent possible errors and biases. In my talk, I will focus on the application of user control in one popular area of AI application, adaptive information access. Adaptive information access systems such as personalized search and recommender systems attempt to model their users to help them in finding the most relevant information. Yet, user modeling and personalization mechanisms might not always work as expected resulting in errors, biases, and suboptimal behavior. Combining the decision power or AI with the ability of the user to guide and control it brings together the strong sides of artificial and human intelligence and could lead to better results. In my talk, I review several projects focused on user control in adaptive information access systems and discuss the benefits and challenges of this approach.
Personalized Online Practice Systems for Learning ProgrammingPeter Brusilovsky
Computer programming is quickly transitioning from being just a key competency in computer and information science majors to being a desired skill for students in a wide range of fields. Yet, it is also one of the most challenging subjects to learn. While learning by doing is a critical component in mastering programming skills, neither the traditional educational process nor standard learning support tools provide sufficient opportunities for programming practice. In this talk, I will present our research on personalized programming practice systems for Java, Python, and SQL, which attempt to bridge this known gap in learning programming. A programming practice system engages students in practicing programming skills beyond a relatively small number of graded assignments and exams. To support learning by doing, an online practice system offers a range of interactive “smart content” such as program animations, worked examples, and various kinds of programming problems with an automatic assessment. The main challenges for online practice systems are to motivate students to practice and to guide them to the most appropriate smart content given their course goals and knowledge levels. In this talk, I will review a range of AI technologies, such as student modeling, navigation support, social comparison, and content recommendation, which support efficient programming practice. I will also discuss how personalized practice system could support COVID-19-influenced switch to online learning while maintaining an extensive level of feedback expected from an efficient learning process.
UMAP 2019 talk Evaluating Visual Explanations for Similarity-Based Recommenda...Peter Brusilovsky
Tsai, Chun-Hua, and Peter Brusilovsky. 2019. "Evaluating Visual Explanations for Similarity-Based Recommendations: User Perception and Performance." In the 27th ACM Conference on User Modeling, Adaptation and Personalization, UMAP 2019, 22-30. Larnaca, Cyprus: ACM.
Course-Adaptive Content Recommender for Course AuthoringPeter Brusilovsky
Developing online courses is a complex and time-consuming
process that involves organizing a course into a sequence of topics and
allocating the appropriate learning content within each topic. This task
is especially difficult in complex domains like programming, due to the
incremental nature of programming knowledge, where new topics extensively
build upon domain concepts that were introduced in earlier lessons.
In this paper, we propose a course-adaptive content-based recommender
system that assists course authors and instructors in selecting the most
relevant learning material for each course topic. The recommender system
adapts to the deep prerequisite structure of the course as envisioned
by a specific instructor, while unobtrusively deducing that structure from
problem-solving examples that the instructor uses to present course concepts.
We assessed the quality of recommendations and examined several
aspects of the recommendation process by using three datasets collected
from two different courses.While the presented recommender system was
built for the domain of introductory programming, our course-adaptive
recommendation approach could be used in a variety of other domains.
Data-Driven Education: Using Big Educational Data to Improve Teaching and Learning. Keynote slides for 15th International Conference on Web-Based Learning, ICWL 2016, Rome, Italy, October 26–29.
From Expert-Driven to Data-Driven Adaptive LearningPeter Brusilovsky
Keynote slides for the Workshop on Advancing Education with Data at the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Halifax, NS, Canada, Aug 14, 2017
Stereotype Modeling for Problem-Solving Performance Predictions in MOOCs and ...Peter Brusilovsky
Proceedings of the 25th Conference on User Modeling, Adaptation and Personalization, UMAP2017, pp 76-84
Stereotypes are frequently used in real life to classify students according to their performance in class. In literature, we can find many references to weaker students, fast learners, struggling students, etc. Given the lack of detailed data about students, these or other kinds of stereotypes could be potentially used for user modeling and personalization in the educational context. Recent research in MOOC context demonstrated that data-driven learner stereotypes could work well for detecting and preventing student dropouts. In this paper, we are exploring the application of stereotype-based modeling to a more challenging task -- predicting student problem-solving and learning in two programming courses and two MOOCs. We explore traditional stereotypes based on readily available factors like gender or education level as well as some advanced data-driven approaches to group students based on their problem-solving behavior. Each of the approaches to form student stereotype cohorts is validated by comparing models of student learning: do students in different groups learn differently? In the search for the stereotypes that could be used for adaptation, the paper examines ten approaches. We compare the performance of these approaches and draw conclusions for future research.
Stereotype Modeling for Problem-Solving Performance Predictions in MOOCs and ...
Encouraging Online Student Reading with Social Visualization
1. Reading
Circle
Encouraging
Online
Student
Reading
with
Social
Visualiza8on
J.
Guerra,
D.Parra,
P.Brusilovsky
School
of
Informa8on
Sciences
University
of
PiGsburgh
ISLG
2013
Workshop,
AIEd
2013.
Memphis,
TN,
USA.
Jul
9th
2013
2. Outline
• Broader
Context
• Goal
• Founda8onal
Concepts
• Our
Approach:
reading
Circle
• Demo:
Reading
Circle
Prototype
in
Ac8on
– Visualiza8on
– User
Interac8on
3. Broader
Context
• Learning
in
groups
vs
CSCL?
• Social
Naviga8on
and
Social
Informa8on
Access
– Using
group
informa8on
access
behavior
to
guide
individual
users
to
the
right
content
• Two
streams
of
work
that
contributed
to
the
Reader
Circle
– Social
naviga8on
for
reading
with
Knowledge
Sea
– Social
guidance
for
problems
in
Progressor
5. Knowledge
Sea
• Focused
on
group-‐level
guidance
– Students
do
follow
visual
cues
– Social
visual
cues
help
to
discover
important
pages
– Much
beGer
guidance
to
important
pages
• Discovered
sizeable
increase
of
mo8va8ons
– Students
access
the
system
more
and
do
more
readings
– With
annota8on-‐based
cues
they
do
more
annota8ons
7. Progressor
• Focused
on
both
guidance
and
engagement
– Social
guidance
on
the
level
of
topics
is
as
good
as
knowledge-‐based
guidance
– Engagement
is
higher
with
social
guidance
• Lessons
learned
– The
ability
to
see
peer
progress
is
at
least
as
important
as
group
progress
– The
interface
organiza8on
has
to
follow
the
student
percep8on
of
content/course
structure
9. Reading
Circle:
Goal
To
encourage
[college]
students’
online
reading
by
developing
a
social
visualiza8on
of
students’
reading
progress
10. Founda8onal
Concepts
• Open
Student
Model:
An
ability
to
view
and
explore
own
progress/knowledge
help
to
reflect
and
plan
• Social
Naviga4on:
“Footprints“
that
can
lead
other
people
to
make
decisions
or
perform
tasks
• Social
Comparison
Theory:
People
tend
to
compare
to
peers.
We
use
it
to
reinforce
produc8ve
behavior
11. Our
Approach:
Reading
Circle
Social
Naviga8on
Reading
and
Annota8on
Panel
Student’s
Progress
13. ISD
Spring
2013
Survey
• Prototype
implemented
and
used
in
graduate
HCIclass
• Introduced
as
an
“addi8onal”
reading
interface
afer
the
first
5
weeks
of
class
• Survey’s
results:
– 48
students
– 36
students
answered
the
survey
– 11
students
used
the
plakorm,
– One
power
user
joined
early,
read
960
pages
with
RC
– 10
“regular”
students
joined
at
different
stages
reading
72
pages
in
average
(pages
LOADED),
with
S.D.
=
67.09
(min=1,
max=198)
14. Survey
results
Statement
(
1
:
Strongly
agree
/
5:
Strongly
disagree)
Avg
Agreem.
*
Avg
Agreem.
It
is
very
important
for
me
to
visualize
my
reading
ac6vity
2.06
1.89
the
reader
visualiza6on
of
my
progress
is
a
good
approach
to
show
my
reading
progress
2.14
1.44
It
is
very
useful
for
me
to
see
reading
progress
or
other
students
in
my
class.
2.36
2.67
It
is
very
useful
for
me
to
see
the
average
class
reading
progress
and
compare
with
my
progress
2.14
2
The
approach
for
class
and
peer
progress
visualiza6on
in
the
reader
applica6on
is
clear
-‐
it
and
allowed
me
to
understand
the
progress
or
other
students
and
compare
it
with
my
own
2.08
1.78
It
is
very
useful
for
me
to
compare
my
current
progress
against
past
weeks
progress
2.42
2.33
The
approach
used
in
the
reader
applica6on
for
showing
my
past
progress
is
clear
and
allowed
me
to
have
a
be>er
picture
of
my
reading
progress
1.97
1.78
15. Discussion
• Good
success
for
the
first
aGempt,
about
¼
of
the
class
used
RC
or
switched
to
it
• Posi8ve
feedback
• Comparison
with
peers
not
working
well
with
a
thumbnail
view
– Upda8ng
pair
comparison
• A
challenge
to
see
the
whole
picture
for
a
teacher
– Crea8ng
a
“Table
Lens”
class
view
16. Thanks!
PAWS
Lab
Want
to
know
more
about
us?
hGp://adapt2.sis.piG.edu/wiki/
peterb@piG.edu