What is ADAM, TeddyBot’s Open-source Research Framework? 

The Future of Educational Technology

The Adaptive Development and Assessment Module (ADAM) is an open-source software framework that (1) enables continual personalization of TeddyBot’s interactions and (2) facilitates replicable educational research and A/B testing of learning interventions and modules. ADAM supports the evolution of educational research, and enables us to continually refine TeddyBot’s learning approaches based on actual, real world data. As an open-source framework, it not only improves TeddyBot’s responses to individual learning patterns but also encourages collaborative development and content diversification.

What are Learning Modules?
Learning Modules are deep-dives into specific learning areas (e.g., archeology, new languages, native plants of the Southwest US, baking, etc,) that complement our subscription learning interactions. They are reviewed and approved by review boards of educators. Their efficacy can be tested through ADAM.

Educator-refined Personalization

ADAM is essential because it enables personalized teaching and it enables educators and researchers to actually test the efficacy of TeddyBot’s learning interventions. For example, researchers can assess whether or not children learn more effectively when counting their own toys versus counting random household or classroom objects. In the future, TeddyBot’s own teaching techniques will prioritize learning interventions that have been shown to be effective for each individual child.

We currently have two school partnerships underway with Riverdale Country School and The Nueva School (see our School Partnerships post for more).

Privacy & Security

Initially, ADAM will be only populated by TeddyBot data. The data will automatically meet the ethical standards of ADAM: anonymized, sensitive data will be filtered out, users own their own data and explicitly consent to each specific usage of their data (e.g., consent for x sequence to be sent to researchers for y study), and the chain of evidence documents the data journey. 

Important Note: TeddyBot does not keep the raw censor data it collects, including video or audio. After the raw data has been processed, the user can consent to store various analytics, patterns, and an abstracted summary. The raw data is automatically deleted.

In the long-term, any product that meets ADAM’s standards can upload data, expanding the scope of research. The framework for exchanging research data ensures secure data handling and ethical usage, maintaining user trust and compliance with global data protection standards.

ADAM Specifications
– Anonymized data
– Unnecessary data filtered out
– Removal of sensitive data
– Users own their own data
– Explicit consent for usage
– Chain of custody

Specifics of ADAM’s Personalization Capabilities

Adaptive Learning Algorithms: The Personalization Powerhouse

At the forefront of ADAM’s capabilities are its adaptive learning algorithms, the cornerstone of personalized education.

  • Dynamic Content Adjustment: Using sophisticated machine learning techniques, ADAM analyzes a child’s interaction patterns, learning pace, and preferences. This analysis allows TeddyBot to dynamically adjust the educational content, ensuring it aligns with each child’s unique learning journey.
  • Continuous Learning Improvement: The adaptive algorithms are designed to evolve, learning from each child’s responses and interactions. This ensures that the educational experience remains relevant, engaging, and effective over time.

Natural Language Processing (NLP): Bridging Communication

Central to ADAM’s functionality is its Natural Language Processing capability. This technology bridges the communication gap between children and TeddyBot, facilitating a more natural and interactive learning experience.

  • Understanding and Response: ADAM’s NLP algorithms are trained to understand children’s speech and text inputs. This allows TeddyBot to not only comprehend queries but also respond in a way that is easily grasped by young learners.
  • Language Skill Enhancement: Through engaging conversations, TeddyBot enhances language skills and cognitive development, making learning both fun and educational.

Emotion Detection: Empathy in Artificial Intelligence

ADAM extends beyond traditional cognitive learning tools by incorporating emotion detection technologies. This feature marks a significant advancement in creating a responsive and empathetic educational environment.

  • Emotion Recognition: By analyzing facial expressions and vocal tones, ADAM detects a range of emotions in real-time. This capability enables TeddyBot to understand and respond to the emotional state of the child. (Reminder: video and audio data are deleted and not stored)
  • Responsive Interactions: Drawing from this emotional insight, TeddyBot adapts its interactions, offering comfort, encouragement, or excitement, thus supporting a child’s emotional and social development.

Real-Time Analytics and Feedback: The Cycle of Improvement

A crucial aspect of ADAM’s functionality is its emphasis on real-time analytics and feedback.

  • Insightful Analytics: ADAM collects and analyzes interaction data, providing valuable insights into each child’s learning progress and engagement levels.
  • Adaptive Teaching Strategies: Based on this real-time feedback, TeddyBot continuously refines its teaching methods and educational content, ensuring an ever-improving and effective learning experience.

Tags:

Comments are closed

Latest Comments
No comments to show.