Intelligent Textbooks

   Sergey Sosnovsky1, Peter Brusilovsky2, Andrew S. Lan3, Richard G. Baraniuk 4
          1
            Utrecht University, Princetonplein 5, Utrecht 3584 CC, the Netherlands
                                  s.a.sosnovsky@uu.nl
      2
        University of Pittsburgh, 135 North Bellefield Ave., Pittsburgh, PA. 15260, USA
                                     peterb@pitt.edu
   3
     University of Massachutsetts Amherst, 140 Governors Dr., Amherst, MA 01003, USA
                               andrewlan@cs.umass.edu
                4
                  Rice University, 6100 Main Street, Houston, TX 77005, USA
                                     richb@rice.edu

Textbooks remain one of the main methods of instruction, but – just like other
educational tools – they have been evolving over the last several decades in many
aspects (how they are created, published, formatted, accessed, and maintained). Most
textbooks these days have digital versions and can be accessed online. Plenty of
textbooks (and similar instructional texts, such as tutorials) are freely available as open
educational resources (OERs). Many commercial textbooks come with libraries of
supplementary educational resources or even distributed as parts of online educational
services built on top of them. The transition of textbooks from printed copies to digital
and online formats has facilitated numerous attempts to enrich them with various kinds
of interactive functionalities including search and annotation, interactive content
modules, automated assessments and more.
   As a result of these enrichments, new research challenges and opportunities emerge
that call for the application of artificial intelligence (AI) methods to enhance digital
textbooks and learners’ interaction with them. Intelligent digital textbooks have the
potential to significantly enhance the online learning experience, the importance of
which is highlighted by the COVID-19 pandemic. There are many research questions
associated with this new area of research; examples include:

• How can one facilitate the access to textbooks and improve the reading process?
• How can one process textbook content to infer knowledge underlying the text and
  use it to improve learning support?
• How can one process increasingly more detailed logs of students interacting with
  digital textbooks and extract insights on learning?
• How can one find and retrieve relevant content “in the wild”, i.e., on the web, that
  can enrich the textbooks?
• How can one better understand both textbooks and student behaviors as they learn
  within the textbook and create personalized learner experiences?
Our workshop series seeks research contributions addressing these and other research
questions related to the idea of intelligent textbooks. While the pioneer work on various
kinds of intelligent textbook technologies has already begun, research in this area is
still rare and spread over several different fields, including AI, human-computer
                                                                                     2

interaction, information retrieval, intelligent tutoring systems, and user modeling. We
hope that this workshop brings together researchers working on different aspects of
intelligent textbook technologies in these fields and beyond to establish intelligent
textbooks as a new, interdisciplinary research field.

The 2021 version of the workshop will build upon the success of the 1st and 2nd
workshops on Intelligent Textbooks that we organized in conjunction with AIED’2019
in Chicago and AIED’2020 online. We intend to further develop this series. Therefore,
we aim at gathering researchers from a wide range of communities that are interested
in all aspects of intelligent textbooks. The 2021 workshop themes include but are not
limited to:
   a) Modelling and representation of textbooks: examining the prerequisite and
        semantic structure of textbooks to enhance their readability;
   b) Analysis and mining of textbook usage logs: analyzing the patterns of learners’
        use of textbooks to obtain insights on learning and the pedagogical value of
        textbook content;
   c) Collaborative technologies: building and deploying social components of digital
        textbooks that enable learners to interact with not only content but other
        learners;
   d) Generation, manipulation, and presentation: exploring and testing different
        formats and forms of textbook content to find the most effective means of
        presenting different knowledge;
   e) Assessment and personalization: developing methods that can generate
        assessments and enhance textbooks with adaptive support to meet the needs of
        every learner using the textbook;
   f) Content curation and enrichment: sorting through external resources on the web
        and finding the relevant resources to augment the textbook and provide
        additional information for learners.

While we did not receive submissions addressing all of these topics, the number of
submitted papers was sufficiently large and the diversity of topics was more than
enough to represent the emerging field as a whole. After a thorough reviewing, where
each submission was reviewed by two to four members of the reviewing committee,
which included workshop organizers and PC members, we selected seven papers for
long presentation and two papers for the short presentations. Additionally, we selected
four papers for interactive demo presentations at the workshop. In the workshop
program, we combined the papers (not including the demos) into three sessions.

Overall, the focus of the workshop was slightly different from the previous year. The
program of the 2019 workshop focused more on such HCI issues as adaptation, and
interactivity. The program of the 2020 workshop focused more on (semi)automated
analysis of textbook content and structure, extraction of knowledge, and integration of
textbooks with other educational systems. This year, the presented papers and demos
covered a broad range of topics related to intelligent textbooks - from approaches to
construct intelligent textbooks to studies based on data collected from modern
textbook platforms. The most popular group of topics this year was related to "smart"
                                                                                       3

use of textbook content to provide additional functionalities, such as automatic
generation of questions and contextual definitions or a textbook-driven chatbot. The
most popular domains were medical textbooks (including dental) and computer
science textbooks.

We want to thank the program committee members that helped us in preparation of
the workshop program:

- Brendan Flanagan
- Atsushi Shimada
- Roger Nkambou
- Vinay Chaudhri
- Paulo Carvalho,
- Noboru Matsuda,
- Cliff Shaffer,
- Erin Walker,
- Elena Glassman,
- Reva Freedman,
- Julio Guerra,
- Ilaria Torre,
- Benjamin Paaßen,
- Paul Denny,
- Zichao Wang,
- Andrew Olney,
- Isaac Alpizar Chacon,
- Debshila Basu Malick.