Total: 26
As AI becomes more widely used across a variety of disciplines, it is increasingly important to teach AI concepts to K-12 students in order to prepare them for an AI-driven future workforce. Hence, educators and researchers have been working to develop curricula that make these concepts accessible to K-12 students. We are designing and developing a comprehensive AI curriculum delivered through a series of carefully crafted activities in an adapted \emph{Snap!} environment for middle-grade students. In this work, we lay out the proposed content of our curriculum and present the design, development, and implementation results of the first unit of our curriculum that focuses on teaching the breadth-first search algorithm. The activities in this unit have been revised after being piloted with a single high-school student. These activities were further refined after a group of K-12 teachers examined and critiqued them during a two-week professional development workshop. Our teachers created a lesson plan around the activities and implemented that lesson in a summer workshop with 14 middle school students. Our results demonstrated that our activities were successful in helping many of the students in understanding and implementing the algorithm through block-based programming while extra supplementary material was needed to assist some other students. In this paper, we explain our curriculum and technology, the results of implementing the first unit of our curriculum in a summer camp, and lessons learned for future developments.
We develop a Multimodal Spatiotemporal Neural Fusion network for MTL (MSNF-MTCL) to predict 5 important students' retention risks: future dropout, next semester dropout, type of dropout, duration of dropout and cause of dropout. First, we develop a general purpose multi-modal neural fusion network model MSNF for learning students' academic information representation by fusing spatial and temporal unstructured advising notes with spatiotemporal structured data. MSNF combines a Bidirectional Encoder Representations from Transformers (BERT)-based document embedding framework to represent each advising note, Long-Short Term Memory (LSTM) network to model temporal advising note embeddings, LSTM network to model students' temporal performance variables and students' static demographics altogether. The final fused representation from MSNF has been utilized on a Multi-Task Cascade Learning (MTCL) model towards building MSNF-MTCL for predicting 5 student retention risks. We evaluate MSNF-MTCL on a large educational database consists of 36,445 college students over 18 years period of time that provides promising performances comparing with the nearest state-of-art models. Additionally, we test the fairness of such model given the existence of biases.
Historically, female students have shown low interest in the field of computer science. Previous computer science curricula have failed to address the lack of female-centered computer science activities, such as socially relevant and real-life applications. Our new summer camp curriculum introduces the topics of artificial intelligence (AI), machine learning (ML) and other real-world subjects to engage high school girls in computing by connecting lessons to relevant and cutting edge technologies. Topics range from social media bots, sentiment of natural language in different media, and the role of AI in criminal justice, and focus on programming activities in the NetsBlox and Python programming languages. Summer camp teachers were prepared in a week-long pedagogy and peer-teaching centered professional development program where they concurrently learned and practiced teaching the curriculum to one another. Then, pairs of teachers led students in learning through hands-on AI and ML activities in a half-day, two-week summer camp. In this paper, we discuss the curriculum development and implementation, as well as survey feedback from both teachers and students.
Dial locks are commonly used to secure a person’s items. Commercially available dial locks often use four or five wheels of letters, allowing a user to select a word as a combination. In order to evaluate the security of these locks, we create a game, with an instance created by the lock designer, and played by a lock owner and a thief. In the game, the lock owner chooses a word as a combination, and the thief creates a brute force strategy to try all possible combinations that yield words until the combination is found. To accomplish the task, the thief will solve a version of the Probabilistic Travelling Salesman Problem (PTSP) by creating an a priori tour through all the words a lock can create. The goal for the game designer, then, is to create a lock configuration that maximizes the expected length of the best possible PTSP tour. This paper describes a Genetic Algorithm (GA) approach to design a near-optimal game, i.e. a lock configuration that makes it as difficult for the thief to crack. An analysis of the output of the GA shows that the locks that the system creates are significantly more secure than both commercial locks, in the context of this game.
Word embeddings, which represent words as dense feature vectors, are widely used in natural language processing. In their seminal paper on word2vec, Mikolov and colleagues showed that a feature space created by training a word prediction network on a large text corpus will encode semantic information that supports analogy by vector arithmetic, e.g., "king" minus "man" plus "woman" equals "queen". To help novices appreciate this idea, people have sought effective graphical representations of word embeddings. We describe a new interactive tool for visually exploring word embeddings. Our tool allows users to define semantic dimensions by specifying opposed word pairs, e.g., gender is defined by pairs such as boy/girl and father/mother, and age by pairs such as father/son and mother/daughter. Words are plotted as points in a zoomable and rotatable 3D space, where the third ”residual” dimension encodes distance from the hyperplane defined by all the opposed word vectors with age and gender subtracted out. Our tool allows users to visualize vector analogies, drawing the vector from “king” to “man” and a parallel vector from “woman” to “king-man+woman”, which is closest to “queen”. Visually browsing the embedding space and experimenting with this tool can make word embeddings more intuitive. We include a series of experiments teachers can use to help K-12 students appreciate the strengths and limitations of this representation.
Offensive words appear in Wordwheel-type puzzles with a high frequency. Previous approaches to eliminating these words have focused largely on eliminating puzzles that might give rise to an offensive word. This work presents a fast, heuristic approach to detecting an offensive word within a puzzle. After a preprocessing stage, the detection occurs with a single bitwise operation on a 64-bit word. Tests show that as long as there are at least 3 taboo words possible in a puzzle, the heuristic approach is faster than a depth-first search of the puzzle. In addition to being fast, the approach is guaranteed to detect all offensive words, and has a low false positive rate.
Auto battlers are a recent genre of online deck-building games where players choose and arrange cards that then compete against other players' cards in fully-automated battles. As in other deck-building games, such as trading card games, designers must balance the cards to permit a wide variety of competitive strategies. We present Ludus, a framework that combines automated playtesting with global search to optimize parameters for each card that will assist designers in balancing new content. We develop a sampling-based approximation to reduce the playtesting needed during optimization. To guide the global search, we define metrics characterizing the health of the metagame and explore their impacts on the results of the optimization process. Our research focuses on an auto battler game we designed for AI research, but our approach is applicable to other auto battler games.
Use of technology-enhanced education and online learning systems has become more popular, especially after COVID-19. These systems capture a rich array of data as students interact with them. Predicting student performance is an essential part of technology-enhanced education systems to enable the generation of hints and provide recommendations to students. Typically, this is done through use of data on student interactions with questions without utilizing important data on the temporal ordering of students’ other interaction behavior, (e.g., reading, video watching). In this paper, we hypothesize that to predict students’ question performance, it is necessary to (i) consider other learning activities beyond question-answering and (ii) understand how these activities are related to question-solving behavior. We collected middle school physical science students’ data within a K12 reading platform, Actively Learn. This platform provides reading-support to students and collects trace data on their use of the system. We propose a transformer-based model to predict students' question scores utilizing question interaction and reading-related behaviors. Our findings show that integrating question attempts and reading-related behaviors results in better predictive power compared to using only question attempt features. The interpretable visualization of the transformer’s attention can be helpful for teachers to make tailored interventions in students’ learning.
In the popular card game Dominion, the configuration of game elements greatly affects the experience for players. If one were redesigning Dominion, therefore, it may be useful to identify game elements that reduce the number of viable strategies in any given game configuration - i.e. elements that are unbalanced. In this paper, we propose an approach that assigns credit to the outcome of an episode to individual elements. Our approach uses statistical analysis to learn the interactions and dependencies between game elements. This learned knowledge is used to recommend elements to game designers for further consideration. Designers may then choose to modify the recommended elements with the goal of increasing the number of viable strategies.
We present an interactive online workshop for K-12 students, which aims in familiarizing students with core concepts of AI. The workshop consists of a variety of resources, inspired by inquiry-based learning techniques, of which we present in detail one module, centered around a browser-based game called Gradient Descent. This module introduces the mathematical concepts behind a gradient descent-based optimization algorithm through the computer game of a treasure hunt at an unknown sea surface landscape. Finally, we report on student feedback for the module in a series of content and language integrated learning in German (CLiLiG) workshops for students aged 14-17 in 30 countries.
This paper presents a structured activity based on a game design to introduce k-12 students in the topic of super-vised machine learning from a practical perspective. The activity has been developed in the scope of an Erasmus+ project called AI+, which aims to develop an AI curriculum for high school students. As established in the AI+ principles, all the teaching activities are based on the use of the student's smartphone as the core element to intro-duce an applied approach to AI in classes. In this case, a smartphone-based game app is developed by students that includes a neural network model obtained with the "Personal Image Classifier" tool of the MIT App Inventor software. From a didactic perspective, the students dealt with supervised learning to solve a problem of image classification. The main learning outcome is the under-standing of how relevant is to develop a reliable machine learning model when dealing with real world applications. This activity was tested during 2021 with more than 50 students belonging to six schools across Europe, all of them enrolled in the AI+ project.
Teaching artificial intelligence (AI) is challenging. It is a fast moving field and therefore difficult to keep people updated with the state-of-the-art. Educational offerings for students are ever increasing, beyond university degree programs where AI education traditionally lay. In this paper, we present an experience report of teaching an AI course to business executives in the United Arab Emirates (UAE). Rather than focusing only on theoretical and technical aspects, we developed a course that teaches AI with a view to enabling students to understand how to incorporate it into existing business processes. We present an overview of our course, curriculum and teaching methods, and we discuss our reflections on teaching adult learners, and to students in the UAE.
Growing awareness of both the demand for artificial intelligence (AI) expertise and of the societal impacts of AI systems has led to calls to integrate learning of ethics alongside learning of technical skills in AI courses and pathways. In this paper, we discuss our experiences developing and piloting the TechHive:AI curriculum for high school youth that integrates AI ethics and technical learning. The design of the curriculum was guided by the following pedagogical goals: (1) to respond to the capacity-building need for critical sociotechnical competencies in AI workforce pathways; and (2) to broaden participation in AI pathways through intentional instructional design to center equity in learning experiences. We provide an overview of the 30-hour learning sequence’s instructional design, and our “4D Framework,” which we use as a heuristic to help students conceptualize and inspect AI systems. We then provide a focused description of one of three 8-hour modules that make up the sequence. Finally, we present evidence of promise from an exploratory study of TechHive:AI with a small sample of students, and discuss insights from implementation, including from our use of established resources for AI learning within the learning sequence as well as those created by our team.
In this experience report, we describe an Artificial Intelligence (AI) Methods in Data Science (DS) curriculum and professional development (PD) program designed to prepare high school teachers with AI content knowledge and an understanding of the ethical issues posed by bias in AI to support their integration of AI methods into existing STEM classrooms. The curriculum consists of 5-day units on Data Analytics, Decision trees, Machine Learning, Neural Networks, and Transfer learning that follow a scaffolded learning progression consisting of introductions to concepts grounded in everyday experiences, hands-on activities, interactive web-based tools, and inspecting and modifying the code used to build, train and test AI models within Google Colab notebooks. The participants in the PD program were secondary school teachers from the Southwest and North-east regions of the United States who represented a variety of STEM disciplines: Biology, Chemistry, Physics, Engi-neering, and Mathematics. We share findings on teacher outcomes from the implementation of two one-week PD workshops during the summer of 2021 and share suggestions for improvements provided by teachers. We conclude with a discussion of affordances and challenges encountered in preparing teachers to integrate AI education into disciplinary classrooms.
In this work, we explain the setup for a technical, graduate-level course on Fairness, Accountability, Confidentiality, and Transparency in Artificial Intelligence (FACT-AI) at the University of Amsterdam, which teaches FACT-AI concepts through the lens of reproducibility. The focal point of the course is a group project based on reproducing existing FACT-AI algorithms from top AI conferences and writing a corresponding report. In the first iteration of the course, we created an open source repository with the code implementations from the group projects. In the second iteration, we encouraged students to submit their group projects to the Machine Learning Reproducibility Challenge, resulting in 9 reports from our course being accepted for publication in the ReScience journal. We reflect on our experience teaching the course over two years, where one year coincided with a global pandemic, and propose guidelines for teaching FACT-AI through reproducibility in graduate-level AI study programs. We hope this can be a useful resource for instructors who want to set up similar courses in the future.
Generative Artificial Intelligence (AI) models are a compelling way to introduce K-12 students to AI education using an artistic medium, and hence have drawn attention from K-12 AI educators. Previous Creative AI curricula mainly focus on Generative Adversarial Networks (GANs) while paying less attention to Autoregressive Models, Variational Autoencoders (VAEs), or other generative models, which have since become common in the field of generative AI. VAEs' latent-space structure and interpolation ability could effectively ground the interdisciplinary learning of AI, creative arts, and philosophy. Thus, we designed a lesson to teach high school students about VAEs. We developed a web-based game and used Plato's cave, a philosophical metaphor, to introduce how VAEs work. We used a Google Colab notebook for students to re-train VAEs with their hand-written digits to consolidate their understandings. Finally, we guided the exploration of creative VAE tools such as SketchRNN and MusicVAE to draw the connection between what they learned and real-world applications. This paper describes the lesson design and shares insights from the pilot studies with 22 students. We found that our approach was effective in teaching students about a novel AI concept.
Intelligent Tutoring Systems have become critically important in future learning environments. Knowledge Tracing (KT) is a crucial part of that system. It is about inferring the skill mastery of students and predicting their performance to adjust the curriculum accordingly. Deep Learning based models like Deep Knowledge Tracing (DKT) and Dynamic Key-Value Memory Network (DKVMN) have shown significant predictive performance compared with traditional models like Bayesian Knowledge Tracing (BKT) and Performance Factors Analysis (PFA). However, it is difficult to extract psychologically meaningful explanations from the tens of thousands of parameters in neural networks, that would relate to cognitive theory. There are several ways to achieve high accuracy in student performance prediction but diagnostic and prognostic reasonings are more critical in learning science. In this work, we present Interpretable Knowledge Tracing (IKT), a simple model that relies on three meaningful features: individual skill mastery, ability profile (learning transfer across skills) and problem difficulty by using data mining techniques. IKT’s prediction of future student performance is made using a Tree Augmented Naive Bayes Classifier (TAN), therefore its predictions are easier to explain than deep learning based student models. IKT also shows better student performance prediction than deep learning based student models without requiring a huge amount of parameters. We conduct ablation studies on each feature to examine their contribution to student performance prediction. Thus, IKT has great potential for providing adaptive and personalized instructions with causal reasoning in real-world educational systems.
In this paper, we introduce a technique for AI generation of the Bullets puzzle, a paper-and-pencil variant of Minesweeper. Whereas traditional Minesweeper can be lost due to the need to guess mine or non-mine positions, our puzzle is fully deducible from a minimal clue set. Puzzle generation is based on analysis and optimization of solutions from a human-like reasoning engine that classifies types of deductions. Additionally, we provide insights to subjective puzzle quality, minimal clue sampling trade-offs, and optimal bullet density.
Automated question quality rating (AQQR) aims to evaluate question quality through computational means, thereby addressing emerging challenges in online learnersourced question repositories. Existing methods for AQQR rely solely on explicitly-defined criteria such as readability and word count, while not fully utilising the power of state-of-the-art deep-learning techniques. We propose DeepQR, a novel neural-network model for AQQR that is trained using multiple-choice-question (MCQ) datasets collected from PeerWise, a widely-used learnersourcing platform. Along with designing DeepQR, we investigate models based on explicitly-defined features, or semantic features, or both. We also introduce a self-attention mechanism to capture semantic correlations between MCQ components, and a contrastive-learning approach to acquire question representations using quality ratings. Extensive experiments on datasets collected from eight university-level courses illustrate that DeepQR has superior performance over six comparative models.
Automated Scoring (AS), the natural language processing task of scoring essays and speeches in an educational testing setting, is growing in popularity and being deployed across contexts from government examinations to companies providing language proficiency services. However, existing systems either forgo human raters entirely, thus harming the reliability of the test, or score every response by both human and machine thereby increasing costs. We target the spectrum of possible solutions in between, making use of both humans and machines to provide a higher quality test while keeping costs reasonable to democratize access to AS. In this work, we propose a combination of the existing paradigms, sampling responses to be scored by humans intelligently. We propose reward sampling and observe significant gains in accuracy (19.80% increase on average) and quadratic weighted kappa (QWK) (25.60% on average) with a relatively small human budget (30% samples) using our proposed sampling. The accuracy increase observed using standard random and importance sampling baselines are 8.6% and 12.2% respectively. Furthermore, we demonstrate the system's model agnostic nature by measuring its performance on a variety of models currently deployed in an AS setting as well as pseudo models. Finally, we propose an algorithm to estimate the accuracy/QWK with statistical guarantees (Our code is available at https://git.io/J1IOy).
Fun, as a game trait, is challenging to evaluate. Previous research explores game arc and game refinement to improve the quality of games. Fun, for some players, is having an even chance to win while executing their strategy. To explore this, we build boards for the game Ticket to Ride while optimizing for a given win rate between four AI agents. These agents execute popular strategies human players use: one-step thinking, long route exploitation, route focus, and destination hungry strategies. We create the underlying graph of a map by connecting several planar bipartite graphs. To build the map, we use a multiple phase design, with each phase implementing several simplified Monte Carlo Tree Search components. Within a phase, the components communicate with each other passively. The experiments show that the proposed approach results in improvements over randomly generated graphs and maps.
In response to the trend that artificial intelligence (AI) is becoming the main driver for social and economic development, enhancing the readiness of learners in AI is significant and important. The state council and the ministry of education of China put AI education for K-12 schools on a high priority in order to foster local AI talents and reduce educational disparities. However, the AI knowledge and technical skills are still limited for not only students but also the school teachers. Furthermore, many local schools in China, especially in the rural areas, are lack of the necessary software and hardware for teaching AI. Hence, we designed and implemented a structured series of AI courses, built on an online block-based visual programming platform. The AI courses are free and easily accessible for all. We have conducted the experimental classes in a local school and collected the results. The results show that the learners in general gained significant learning progress on AI knowledge comprehension, aroused strong interests in AI, and increased the degree of satisfaction towards the course. Especially, our practices significantly increased computational thinking of the students who were initially staying at a lower level.
The Hands-On AI Projects for the Classroom series, a collection of five guides, includes interactive projects that can be used by teachers across grade levels and subject areas to teach K-12 students about artificial intelligence (AI).
The StoryQ environment provides an intuitive graphical user interface for middle and high school students to create features from unstructured text data and train and test classification models using logistic regression. StoryQ runs in a web browser, is free and requires no installation. AI concepts addressed include: features, weights, accuracy, training, bias, error analysis and cross validation. Using the software in conjunction with curriculum currently under development is expected to lead to student understanding of machine learning concepts and workflow; developing the ability to use domain knowledge and basic linguistics to identify, create, analyze, and evaluate features; becoming aware of and appreciating the roles and responsibilities of AI developers;. This paper will consist of an online demo with a brief video walkthrough.
We will demonstrate blocks integrated into Snap! capable of a wide range of AI services, interactive AI programming guides, and a selection from thirty sample projects. Sessions and workshops in both school settings and informal learning contexts have been held in many countries. The full version of this paper includes descriptions of the Snap! blocks and unpublished descriptions of student experiences in India.