ETEXTBOOKS: What's Their Future and How Can Libraries Prepare?
by Neil Dixon
AI is found in digital services we use every day, such as online banking, social media, and ecommerce. When libraries have adopted AI, it has mainly been in incremental service or procedural improvements, such as help desk chatbots (Cox et al. 2019). The growth of AI and the possibility that administrative library tasks could be automated can lead to fears of change and job security. Given these fears, it is important to educate library staffers about AI to help us anticipate possible changes in the future and adapt accordingly. At Anglia Ruskin University’s (ARU) library, we aimed to do this by running narrative fiction writing workshops on how future etextbooks that use AI might impact libraries. This article describes current etextbook technology, some potential ideas for future etextbook features and how they could impact libraries, and an explanation of the strategic planning workshops we conducted with library staff members.
|This is just one example of the growing trend for academic libraries to work with the university to provide specific services, rather than the library being a gatekeeper of a collection or space ...
What’s the difference between an etextbook and a library ebook? An etextbook is a digitized version of the key “text” book reading for a course. In 2019, ARU procured Kortext to provide 5,000 students with the key text for each of their first-year modules on a one-to-one loan model for the year, meaning the students could personalize their reading experience and download the key texts on different digital devices. The introduction of this new technology on such a large scale meant there would be some changes to how the library supported students and staffers, which is why we chose to specifically investigate the future of etextbooks.
When users log in to the Kortext etextbook platform, they are greeted by a bookshelf displaying covers from the print editions of their allocated key texts, with an indication of the format (EPUB or PDF). Users can select which title to read from there or search via keyword to locate chapters, author references, or topics across all of their key texts. The Kortext reader offers functions such as bookmarking, exportable notes and highlights, customizable fonts, text sizes and colors, and layout (see Figure 1 for an example). If students have a font preference, they can choose a font such as OpenDyslexic to make it easier for them to read, and these settings are saved in the mobile app or online reader.
Kortext’s built-in learning analytics allow user activity to be checked at an institutional, book title, or user level. Lecturers who use the key text in their teaching, for instance, can get an aggregated overview of the keyword searches, which is a straightforward way to check whether the students understand a concept and are searching for the right words. Lecturers also have access to the platform, so they can show relevant diagrams or charts from a key text in a lecture rather than copying the image onto slides; students could refer to a page on their own device in the same way as if they had a print copy.
When adopted at scale, etextbook platforms can result in a change in how libraries support users, especially students. On one level, this is about libraries promoting the service to make sure all eligible students log in to retrieve their key texts. On another level, to get the most from etextbooks, students and staffers need to be taught how to use the features, despite the perception that young people entering university are Digital Natives and can easily learn how to use technology (Ross et al. 2017). Engaging with lecturers to select specific key texts—and encouraging them to use the platform in their teaching—is another change to library support roles (Tatham and Moore 2019). This can help increase student usage. The extra support required means the library needs to work with the provider to coordinate opportunities for training and solving technical issues, in addition to introducing students to other platforms such as reading lists. The library also needs to collaborate with other support services, such as IT, to secure the technical capacity required to assign all users the required key texts and integrate the platform with institutional technologies such as the virtual learning environment. This is just one example of the growing trend for academic libraries to work with the university to provide specific services, rather than the library being a gatekeeper of a collection or space (Cox 2019; Dempsey and Malpas 2018).
AI and the Future of Etextbooks
Given the changes the introduction of a large-scale etextbook platform brings to libraries, how can etextbook platforms evolve in the future, and how can libraries adapt? In terms of future technological integrations, AI and machine learning present the most promising areas of research for etextbooks because of the opportunity to give a unique and personalized learning experience to the user. AI refers to a non-human intelligence that can solve specific tasks, such as sorting, predicting, or classifying data automatically (Google 2020). In the case of etextbooks, the AI would receive the content of each key text and user’s interactions as an input. Machine learning is a subfield of AI, which applies to computer algorithms that can be trained and improve over time.
In one scenario, it is possible to envisage partially automating the creation of a unique etextbook using preselected material. In 2019, Springer published the first machine-generated textbook. Its topic was lithium-ion batteries, and it was created by an algorithm called Beta Writer (Beta Writer 2019). The process involved automatically sorting, classifying, and organizing the content from hundreds of journal articles and outputting it as an original manuscript. The parameters to organize the chapters were set beforehand, comprising an introduction, short summaries of research within the topic subject headings, a conclusion, and references that link back to the original source. Springer already has plans for more machine-generated etextbooks; as other publishers develop this technology, there will be an exponential increase in the number of publications. There is a continuing role for librarians in educating users on distinguishing among authoritative sources by detecting the authorship and perspective of machine-generated content. In the same way we teach effective searching for information, we will need to educate users on the process an AI utilizes to create a publication, as well as to detect potential copyright issues.
It is also possible to envision offering new ways to interact with an etextbook by using AI. One method could be an intelligent assistant within the etextbook, meaning users could ask a question and receive an answer, both using natural language (Hollingsworth and Narayanan 2016). Currently, the search function within etextbooks is generally employed as a user interface for fact-checking or locating specific areas of the book, and the results are the sections in the text that link to the specific location.
With an intelligent assistant, the search can be a way for users to think through a process or topics in a new light. Talk to Books, for example, displays passages from books in response to natural language questions, which can help with inspiration or insights (Google AI 2020). Talk to Books acts more as a conversational agent, and by presenting a problem, the AI can aid decision making rather than entering facts or figures to make an analysis (Johnson 2018), which is the case in traditional search. Similarly, there is scope for etextbooks to use adaptive learning, taking datapoints (such as quiz scores, notes, or highlights) and changing the content depending on a user’s progress (Hollingsworth and Narayanan 2016).
The integration of AI into etextbooks will bring with it a significant change: the availability of additional learning analytics data. Library staffers will need to be capable of using tools to analyze this data to understand what it means and how they can use it to make informed decisions. At an operational level, the data provides opportunities such as being able to target information literacy support, submitting regular reports to faculties to assist in reading list development, and working with other departments to make the data interoperable with other university systems. In addition, there will be a significant evolution in how libraries procure these platforms, which will involve making sure the data conforms to data protection regulations, whether content is fully accessible given the possibility for adapted learning, and awareness of any future regulations that could govern platform selection.
Preparing Library Staffers for Future Etextbook Technologies
To encourage library staff members to think about how these technologies might impact libraries, we ran workshops with two groups of staffers (both internal and external), in which participants wrote fictional narratives imagining themselves in an alternative future where there were new ways to interact with etextbooks.
Based on a literature review, we created a list of nine features that were an innovative improvement to etextbooks, but all needed to have a proof of concept, rather than being speculation. We categorized the features into content, learning, and platform so we had a broad range. Using Microsoft Publisher, we designed printable cards with a straightforward description of the feature on the front and an explanatory note and journal article reference on the back (see Figure 2 and Figure 3). The workshop format was influenced by Huusko et al. (2018), who used a similar method; ours comprises two activities condensed into 60 minutes.
After a brief introduction, we started a 5-minute exercise in which we asked, “If etextbooks were a building, food, or person, what would that be?” This helped participants get into a creative mindset, and examples such as “A young Harry Potter—good but not fully functional” summed up the consensus on etextbooks during the discussion that followed the exercise. After this, we explained the next activity and asked participants to take three cards from the stacks that we’d placed in the room. The brief was to imagine a future in 5–10 years in which etextbooks with features listed on the cards existed and write a personal narrative of themselves or a student using them. As guidance, we handed out two examples of narrative writing that we had composed: one a review of a fictional future etextbook, the other a narrative from the perspective of a future student. We made it clear that the purpose was to think creatively about how libraries might change and how this might impact their roles, rather than judge the merits of their creative writing.
The narratives were creative and in the spirit of the session, with the participants proposing etextbook features such as speed-reading modes, instant feedback, and greater scope for collaboration. We received positive feedback, with comments such as, “just what I wanted from this session—a chance to think outside the box and tackle a problem with a fresh mind. Thank you!”
Running a workshop like this is a good way to inspire library staffers in a creative and practical way. Not all of the scenarios described in the narratives were probable, or even possible, but by imagining the future, we can prepare for how things might well change.
Beta Writer. Lithium-Ion Batteries: A Machine-Generated Summary of Current Research. Springer, 2019.
Cox, Andrew M., Stephen Pinfield, and Sophie Rutter. “The Intelligent Library.” Library Hi Tech, 2019.
Dempsey, Lorcan and Constance Malpas. “Academic Library Futures in a Diversified University System.” Higher Education in the Era of the Fourth Industrial Revolution, pp. 65–89, 2018.
Google. Machine Learning Glossary, developers.google.com/machine-learning/glossary . Accessed June 27, 2020.
Google AI. Talk to Books, books.google.com/talktobooks . Accessed June 27, 2020.
Huusko, Maria, Yiying Wu, and Virpi Roto. “Structuring and Engaging: The Roles of Design Fictions in a Co-Design Workshop.” In Proceedings of the 30th Australian Conference on Computer-Human Interaction, pp. 234–241, 2018.
Hollingsworth, M.L. and N.H. Narayanan. “Building a Better eTextbook.” Bulletin of the IEEE Technical Committee on Learning Technology 18, No. 2/3, 2016.
Johnson, Ben. “Libraries in the Age of Artificial Intelligence.” Computers in Libraries 38, No. 1, 2018.
Ross, Bella, Ekaterina Pechenkina, Carl Aeschliman, and Anne-Marie Chase. “Print Versus Digital Texts: Understanding the Experimental Research and Challenging the Dichotomies.” Research in Learning Technology 25, 2017.
Society of College, National and University Libraries (SCONUL). About Sconul, sconul.ac.uk/page/about-sconul. Accessed June 27, 2020.
Tatham, Suzanne and Annette Moore. “How Valuable Are E-Textbooks to the Student Experience? An Analysis of E-Textbook Provision at the University of Sussex.” Against the Grain 31, No. 3, pp. 22–24, 2019.