AI Credits
Credits can be used to pay for the computational resources, tools, or services that are offered by an AI provider. For example, credits are used for AI-powered actions in Microsoft 365 apps.
AI in Academic Publishing
Applications of AI to streamline editorial processes, improve manuscript quality through language refinement, and detect data inconsistencies or errors in submissions.
AI-Driven Academic Advising
The use of AI systems to assist students with course selection, career planning, and personalized learning pathways based on their preferences and academic performance.
AI-Enhanced Peer Review
The use of AI systems to assist in the peer review process for academic publications by identifying inconsistencies, suggesting improvements, or detecting potential biases in submissions.
AI-Powered Research Tools
Tools that use AI to assist researchers by summarizing articles, finding relevant citations, organizing literature, or generating hypotheses from existing data.
Adaptive Assessment Systems
Assessment platforms that adjust the difficulty, pacing, or content of questions based on a student’s performance, providing a more personalized evaluation experience.
Adaptive Learning
A method of customizing educational content based on a learner’s performance, adjusting factors like difficulty, pacing, and type of support provided.
Algorithm
A step-by-step procedure or formula used to perform a task or solve a problem, often used in AI and machine learning models.
Artificial General Intelligence (AGI)
A type of AI that is capable of performing any intellectual task that a human can, unlike narrow AI, which is specialized in one task.
Artificial Intelligence (AI)
The simulation of human intelligence processes by machines, especially computer systems. It includes learning, reasoning, problem-solving, perception, and language understanding.
Artificial Narrow Intelligence (ANI)
AI systems that are designed and trained to perform a specific task, such as image recognition or playing chess.
Automated Literature Reviews
AI systems that analyze and synthesize existing research papers to generate summaries, identify gaps, or highlight emerging trends in a field of study.
Autonomous Systems
AI-powered systems capable of performing tasks without human intervention, such as self-driving cars or drones.
Bias in AI
A systematic error in AI models, often caused by biased training data, that leads to unfair or inaccurate results.
Black Boxes
AI systems whose internal decision-making processes are opaque, making it difficult to understand how outcomes are reached. Techniques like interactive machine learning (IML) and explainable AI aim to improve transparency.
Challenge
The explanation may not always be accurate, raising concerns about trust. As a solution, researchers advocate for interpretable models that are inherently transparent.
ChatGPT Models
These are chat-based systems built on neural network transformer models for natural language processing. ChatGPT models (1) generate responses to questions, (2) are pre-trained on extensive written material, and (3) utilize transformer architecture to handle sentence-level context.
Chatbot
A software application used to conduct a conversation via text or voice, powered by AI and NLP technologies.
Collaborative AI Research Platforms
Platforms that integrate AI to enhance collaborative efforts among researchers, enabling better coordination, data sharing, and interdisciplinary projects.
Computer Vision
A field of AI focused on enabling machines to interpret and make decisions based on visual data from the world, such as images or videos.
Convolutional Neural Network (CNN)
A type of deep learning model primarily used for processing and analyzing visual data like images and videos.
Critical AI
An approach that critiques and reflects on existing AI structures, focusing on ethical and historical implications.
Data
Information about people, objects, or environments that AI systems use for training and operation.
Data Augmentation
A technique used to increase the size and diversity of training datasets by creating modified versions of existing data, such as flipping or rotating images or adding noise to text data, to improve model performance.
Data Preprocessing
The process of cleaning, organizing, and transforming raw data into a suitable format for training machine learning models.
Deep Learning
A specialized form of machine learning that uses neural networks with many layers (hence “deep”) to analyze large amounts of data, often applied in tasks like image recognition and natural language processing.
Digital Humanities
An interdisciplinary field where computational methods, including AI, are applied to analyze, interpret, and visualize humanities data, such as historical texts or cultural artifacts.
Edge AI
AI systems that process data locally on edge devices (like smartphones or IoT sensors) rather than relying on cloud computing. This reduces latency and enhances data privacy.
Educational Data Mining (EDM)
A field focused on applying data mining techniques to educational data to discover patterns and insights that can improve teaching and learning.
Equity-Focused AI in Education
AI systems designed to address disparities in education by providing personalized support for underrepresented or disadvantaged students.
Ethical AI
The practice of designing, developing, and deploying AI systems with principles that promote fairness, transparency, accountability, and respect for human rights. Ethical AI aims to mitigate risks such as bias, misuse, or unintended harm.
Explainability
The degree to which a model’s decisions and operations can be understood by humans. Important for transparency and trust in AI systems.
Explainable Machine Learning (XML) or Explainable AI (XAI)
XML/XAI comprises methods and processes that clarify how machine learning algorithms produce results. These approaches help developers ensure systems function correctly and comply with regulatory standards.
Feature Engineering
The process of selecting, modifying, or creating new features from raw data to improve the performance of a machine learning model.
Federated Learning
A collaborative machine learning method where models are trained across multiple devices or servers using local data. This approach enhances privacy since data doesn’t need to be centralized during training.
Few-Shot Learning (FSL)
A machine learning technique where models learn to perform tasks or recognize new categories with minimal labeled examples.
Foundation Models
Pre-trained models on vast datasets that serve as a starting point for developing specialized systems. Their use has raised concerns around data bias and trustworthiness.
Generative AI (GenAI)
GenAI refers to machine learning models that generate content such as text, images, music, videos, or 3D models. ChatGPT is an example of Generative AI.
Generative Adversarial Network (GAN)
A class of machine learning frameworks where two neural networks (a generator and a discriminator) are trained simultaneously to generate data that is indistinguishable from real data.
Gradient Descent
An optimization algorithm used in training machine learning models. It iteratively adjusts the model’s parameters to minimize a loss function, guiding the model toward better predictions.
Hidden Layer(s)
Performs calculations, adjusting connection strengths.
Human-Centered Perspective
This philosophy emphasizes collaboration between AI systems and humans, ensuring that AI augments rather than replaces human abilities, particularly in education.
Input Layer
Receives the data.
Intelligence Augmentation (IA)
IA refers to systems designed to enhance human capabilities, such as automating redundant tasks to free up time for uniquely human activities.
Intelligent Tutoring Systems (ITS)
Digital systems that provide personalized feedback to students using rule-based or machine learning techniques, supporting adaptive learning.
Interpretable Machine Learning (IML)
IML refers to models designed to be transparent, providing explanations for their decisions inherently rather than relying on post-hoc analysis.
Key Distinction
“Explainable” models provide insights after a decision is made, often using secondary algorithms to interpret the “black box” models.
Large Language Models (LLMs)
LLMs are foundation models for GenAI, trained on extensive text datasets using neural networks and deep learning techniques. Examples include OpenAI’s GPTs, Meta’s LLaMA, and Google’s PaLM. LLMs predict text by modeling statistical word patterns.
Learning Analytics
The use of data, statistical methods, and AI tools to analyze learner behavior and improve educational outcomes. Learning analytics helps educators make data-driven decisions about curriculum design, teaching methods, and student support.
MOOCs and AI Integration
Massive Open Online Courses (MOOCs) that utilize AI for adaptive learning, real-time feedback, and personalized course recommendations to enhance learner engagement.
Machine Learning (ML)
A subset of AI that involves training algorithms to recognize patterns in data and make predictions or decisions based on that data.
Model Evaluation
The process of assessing the performance of a machine learning model using various metrics, such as accuracy, precision, recall, F1 score, etc.
Model Fine-Tuning
The process of adapting a pre-trained model to perform a specific task by training it further on a smaller, task-specific dataset. Fine-tuning allows models to achieve high accuracy in specialized applications.
Natural Language Processing (NLP)
A field of AI that focuses on the interaction between computers and human languages, including tasks like text analysis, language generation, and speech recognition.
Neural Architecture
The design of a neural network includes the structure and number of layers, types of neurons, and connections.
Neural Network
A computational model inspired by the human brain, composed of layers of nodes (neurons) that process data. Used in deep learning.
Neural Networks (NN)
Neural networks, also known as artificial neural networks (ANN), are a subset of machine learning algorithms inspired by the human brain’s neural connections. Data flows through layers of nodes.
Neuro-Symbolic AI
An approach that combines neural networks and symbolic reasoning to create AI systems capable of learning and logical problem-solving.
Open Educational Resources (OER) with AI
AI-enhanced platforms that support the creation, curation, and customization of free educational resources for educators and learners worldwide.
Output Layer
Produces results.
Overfitting
A modeling error in machine learning where a model learns the details and noise of the training data to the point it negatively impacts its performance on new data.
Plagiarism Detection Tools
AI-powered systems designed to identify potential instances of plagiarism in academic writing by comparing text to large databases of published works and online content.
Predictive Analytics in Higher Education
The use of AI to analyze historical and real-time data to predict student outcomes, such as retention rates, graduation probabilities, or career success.
Recurrent Neural Network (RNN)
A class of neural networks designed for sequence prediction tasks, where previous outputs are used as inputs for subsequent steps (used in speech recognition, time series prediction).
Reinforcement Learning
A type of machine learning where an agent learns by interacting with an environment and receiving rewards or penalties based on its actions.
Reinforcement Learning (RL)
A type of machine learning where algorithms learn by interacting with an environment and receiving feedback in the form of rewards or penalties. The system adjusts its actions to maximize cumulative rewards over time.
Research Ethics in AI
The application of ethical principles to guide AI research and development. This includes considerations for privacy, informed consent, fairness, and minimizing potential harms in studies involving AI.
Robots
Robots are mechanical devices designed to perform physical tasks for humans. In contrast, “bots” are software agents that perform digital tasks, such as offering help in intelligent tutoring systems. While both can incorporate AI, including machine learning, it’s not a requirement. AI enhances their adaptability and complexity when performing tasks.
Self-Attention Mechanism
A system that determines the most critical parts of input data, inspired by human attention. It enables AI models to focus effectively and encode information contextually.
Semi-Supervised Learning
A hybrid approach combining supervised learning (using labeled data) and unsupervised learning (using unlabeled data) to improve learning efficiency, especially when labeled data is limited.
Singularity
A hypothetical point in the future where AI surpasses human intelligence, leading to rapid, unpredictable advances.
Smart Classrooms
Classrooms equipped with AI-powered tools to automate administrative tasks, provide real-time feedback, and create interactive, immersive learning experiences.
Supervised Learning
A type of machine learning where the model is trained on labeled data (i.e., data that includes both input and corresponding output).
Test Data
The dataset used to evaluate the performance of a trained model, to ensure that it generalizes well to unseen data.
Training Data
The dataset used to train a machine learning model, where the input data is paired with the correct output.
Transfer Learning
A machine learning technique where a model developed for a particular task is reused as the starting point for a model on a second task.
Transformer Models
Transformer models are neural networks used in Generative AI and natural language processing. They employ a self-attention mechanism to focus on essential input features, improving processing and output.
Turing Test
A test proposed by Alan Turing to determine if a machine exhibits human-like intelligence, specifically if it can convince a human that it is human during a conversation.
Underfitting
A situation in machine learning where a model is too simple to capture the underlying patterns in the data, leading to poor performance.
Unsupervised Learning
A type of machine learning where the model is trained on unlabeled data, and it must find patterns and structure on its own.
User Experience Design/User Interface Design (UX/UI)
UX/UI refers to the creation and refinement of user interactions with products or technologies. These design methodologies aim to improve usability and ensure seamless experiences, beyond just AI applications.
Virtual Research Assistants
AI-powered tools or chatbots designed to help researchers with routine tasks, such as literature reviews, data management, or scheduling experiments.
Zero-Shot Learning (ZSL)
A method in machine learning where models can make predictions or classify data in categories they haven’t been explicitly trained on, by leveraging generalizable knowledge.
