SlideShare a Scribd company logo
YAI4Edu: an
Explanatory AI to
Generate
Interactive e-Books
for Education
F. Sovrano, K. Ashley, P. Brusilovsky, F. Vitali
Dipartimento di Informatica


Alma Mater – Università di Bologna
Introduction
• Most of our educational content (e.g. books) is
static: sub-optimal and time consuming for a human
reader.


• We study how to automatically enhance (static)
books making them interactive:


• reducing the sparsity of relevant information;


• increasing the explanatory power of the
medium by linking it to related books.


• We present YAI4Edu: a first prototype of
Explanatory AI to generate interactive e-books for
education.
Background Theory
• We exploit a recent theory of explanations from
Ordinary Language Philosophy


• Explaining: a process of illocutionary question
answering.


• Assumption: the goal of an educational e-book is
to explain something to the reader.


• Idea: organising the explanatory space (the space
of all possible bits of explanation) as clusters of
questions and answers is beneficial for the reader.
Background Theory
• Explaining requires question answering but it is
not equivalent to it.


• Difference between explaining and question
answering: illocution.


• Illocution: an act that involves informed and
pertinent answers not just to the main question,
but also to other (archetypal) questions of various
kinds that are relevant to the explainee.


• Archetypal Question: why, what, when, who,
how, etc.
YAI4Edu
• YAI4Edu relies on AI for: knowledge graph
extraction, question-answer extraction and
answer retrieval.


• YAI4Edu converts a book into a hyper-graph where
information can be either explored through
overviewing or searched through questioning.


• Interaction Features: word glosses (called
overviews) and a special kind of search box that
allows the reader to get answers about any English
question.
eBook
Interactive Elements
Exploration: Overviews
Exploration: QA
Discussion
• Hypothesis: explaining is somehow about question
answering so that the more questions a book can
answer, the more it explains.


• Expectations: The explanatory power of a book
can be improved by making it interactive in a way
that helps its readers to identify the most
important questions to be answered and to get
answers about their own questions.


• Experiment: see whether students acquire new
knowledge more deeply with interactive e-books
generated by YAI4Edu.
www.unibo.it
Speaker: Francesco Sovrano
Dipartimento di Informatica – Scienze e Ingegneria


Alma mater – Università di Bologna
Website: unibo.it/sitoweb/francesco.sovrano2


E-mail: francesco.sovrano2@unibo.it

More Related Content

PPTX
Harnessing Textbooks for High-Quality Labeled Data: An Approach to Automatic ...
PDF
Toward Eliminating Hallucinations: GPT-based Explanatory AI for Intelligent T...
PDF
Layout- and Activity-based Textbook Modeling for Automatic PDF Textbook Extra...
PPTX
Exploring the Content Ecosystem of the First Open-source Adaptive Tutor and i...
PPTX
Advancing Intelligent Textbooks with Automatically Generated Practice: A Larg...
PPTX
Creating Session Data from eTextbook Event Streams
PDF
Augmenting Digital Textbooks with Reusable Smart Learning Content: Solutions ...
PDF
Interactions of reading and assessment activities
Harnessing Textbooks for High-Quality Labeled Data: An Approach to Automatic ...
Toward Eliminating Hallucinations: GPT-based Explanatory AI for Intelligent T...
Layout- and Activity-based Textbook Modeling for Automatic PDF Textbook Extra...
Exploring the Content Ecosystem of the First Open-source Adaptive Tutor and i...
Advancing Intelligent Textbooks with Automatically Generated Practice: A Larg...
Creating Session Data from eTextbook Event Streams
Augmenting Digital Textbooks with Reusable Smart Learning Content: Solutions ...
Interactions of reading and assessment activities

More from Sergey Sosnovsky (20)

PDF
Parallel Construction: A Parallel Corpus Approach for Automatic Question Gene...
PDF
Automatic Question Generation for Evidence-based Online Courseware Engineering
PDF
Reading Comprehension Quiz Generation using Generative Pre-trained Transformers
PPTX
Mathematical Language Processing via Tree Embeddings
PPTX
Contextual Definition Generation
PPTX
Transforming Textbooks into Learning by Doing Environments: An Evaluation of ...
PPTX
Generation of Assessment Questions from Textbooks Enriched with Knowledge Models
PPTX
Using Semantics of Textbook Highlights to Predict Student Comprehension and K...
PPTX
Dental TutorBot: Exploitation of Dental Textbooks for Automated Learning
PDF
What's in a textbook
PPTX
Using Programmed Instruction to Help Students Engage with eTextbook Content
PPTX
Adding Intelligence to a Textbook for Human Anatomy with a Causal Concept Map...
PPTX
Interlingua: Linking Textbooks Across Different Languages
PPTX
Student Modeling with Automatic Knowledge Component Extraction for Adaptive T...
PDF
What’s a Textbook? Envisioning the 21st Century K-12 Text
PPTX
BBookX: Creating Semi-Automated Textbooks to Support Student Learning and Dec...
PPTX
PASTEL: Evidence-based Learning Engineering Method to Create Intelligent Onli...
PPTX
Interactive and Personalized Activity eBooks for Learning to Read: The iRead ...
PDF
An Overview of Recent Developments in Intelligent e-Textbooks and Reading Ana...
PPTX
Scaffolded, Scrutable Open Learner Model (SOLM) as a Foundation for Personali...
Parallel Construction: A Parallel Corpus Approach for Automatic Question Gene...
Automatic Question Generation for Evidence-based Online Courseware Engineering
Reading Comprehension Quiz Generation using Generative Pre-trained Transformers
Mathematical Language Processing via Tree Embeddings
Contextual Definition Generation
Transforming Textbooks into Learning by Doing Environments: An Evaluation of ...
Generation of Assessment Questions from Textbooks Enriched with Knowledge Models
Using Semantics of Textbook Highlights to Predict Student Comprehension and K...
Dental TutorBot: Exploitation of Dental Textbooks for Automated Learning
What's in a textbook
Using Programmed Instruction to Help Students Engage with eTextbook Content
Adding Intelligence to a Textbook for Human Anatomy with a Causal Concept Map...
Interlingua: Linking Textbooks Across Different Languages
Student Modeling with Automatic Knowledge Component Extraction for Adaptive T...
What’s a Textbook? Envisioning the 21st Century K-12 Text
BBookX: Creating Semi-Automated Textbooks to Support Student Learning and Dec...
PASTEL: Evidence-based Learning Engineering Method to Create Intelligent Onli...
Interactive and Personalized Activity eBooks for Learning to Read: The iRead ...
An Overview of Recent Developments in Intelligent e-Textbooks and Reading Ana...
Scaffolded, Scrutable Open Learner Model (SOLM) as a Foundation for Personali...

YAI4Edu: an Explanatory AI to Generate Interactive e-Books for Education

  • 1. YAI4Edu: an Explanatory AI to Generate Interactive e-Books for Education F. Sovrano, K. Ashley, P. Brusilovsky, F. Vitali Dipartimento di Informatica Alma Mater – Università di Bologna
  • 2. Introduction • Most of our educational content (e.g. books) is static: sub-optimal and time consuming for a human reader. • We study how to automatically enhance (static) books making them interactive: • reducing the sparsity of relevant information; • increasing the explanatory power of the medium by linking it to related books. • We present YAI4Edu: a first prototype of Explanatory AI to generate interactive e-books for education.
  • 3. Background Theory • We exploit a recent theory of explanations from Ordinary Language Philosophy • Explaining: a process of illocutionary question answering. • Assumption: the goal of an educational e-book is to explain something to the reader. • Idea: organising the explanatory space (the space of all possible bits of explanation) as clusters of questions and answers is beneficial for the reader.
  • 4. Background Theory • Explaining requires question answering but it is not equivalent to it. • Difference between explaining and question answering: illocution. • Illocution: an act that involves informed and pertinent answers not just to the main question, but also to other (archetypal) questions of various kinds that are relevant to the explainee. • Archetypal Question: why, what, when, who, how, etc.
  • 5. YAI4Edu • YAI4Edu relies on AI for: knowledge graph extraction, question-answer extraction and answer retrieval. • YAI4Edu converts a book into a hyper-graph where information can be either explored through overviewing or searched through questioning. • Interaction Features: word glosses (called overviews) and a special kind of search box that allows the reader to get answers about any English question.
  • 9. Discussion • Hypothesis: explaining is somehow about question answering so that the more questions a book can answer, the more it explains. • Expectations: The explanatory power of a book can be improved by making it interactive in a way that helps its readers to identify the most important questions to be answered and to get answers about their own questions. • Experiment: see whether students acquire new knowledge more deeply with interactive e-books generated by YAI4Edu.
  • 10. www.unibo.it Speaker: Francesco Sovrano Dipartimento di Informatica – Scienze e Ingegneria Alma mater – Università di Bologna Website: unibo.it/sitoweb/francesco.sovrano2 E-mail: francesco.sovrano2@unibo.it