An established UK publisher of educational textbooks approached Fountech Solutions with the idea of being able to digitise their content, so that they could be accessed via subscription by educators, institutions and individual learners.
Crucially, the client wanted to enable an interactive experience, where ‘readers’ could ask questions about specific subjects within an e-book, find particular textual passages very easily, and have the AI driven software generally act asa ‘Q&A’ and research tool in addition to simple cataloguing functions. The client further realised that such a system could have enormous implications for licensing software within the educational technology market.
E-books aren’t currently interactive. They are often manually ‘tagged’ by publishers using limited taxonomy. Naturally, if one wished to purchase an e-book about, say, the history of WW1, a GoogleBooks search or academic resource would reveal a variety of paper and e-books by subject, category, author, date etc. The Dewey System, an analogue decimal classification system from the 19th century, is still used today.
But the client wanted much more than cataloguing facilities. Static e-books, which can be accessed by an e-reader of some kind, can be searched just by keyword or manually applied tag.
Our client wanted to find a way of unlocking much further potential to put the e-book into an interactive, insightful, engagement platform.
First of all AI had to be created to learn about the book, using algorithms:
Subject classification. Categorising chapters, for example, into one of many categories. This is especially useful for multi-themed books such as educational textbooks or 'short story’ anthologies. Classification can also assign type and genre (e.g. fiction, horror).
Chapters can refer to similar content or expand upon previous content. The AI algorithms can create connections between chapters of such content and create/suggest learning sets within the smart textbook.
Some specific names, professions, physical locations etc etc are extracted from text. These can be used to conduct a broader search, external to the book itself.This can be achieved either ‘a priori’ or 'on the fly’. Sources can be existing references from any public domain or other materials from the same publisher. For example, a particular chapter of a history book might refer to an historic character, in which case, the reader could click on that character's name or ask by voice to learn more about them.
The contents and meta-data are stored in a graph-based knowledge-representation system, which is used to easily retrieve interrelated data.
Chapter summaries. Each chapter is summarized to a synopsis of chosen length, say 400 words per chapter. Each chapter can be categorised by themes.Moreover, interrelated information can be asked of the system e.g. "Can you tell me about the central characters of this book and where they first appear?"
Readers may be bilingual but may still feel more comfortable having material presented in their native language.The platform allows people to choose book segments to translate into a different language. This can help with their understanding and engagement analysis efforts, not least if they are using the SmartTextbook in order to learn a new language!
The rate of consumption of a smart textbook can help to assess a reader’s understanding, in terms of them asking for repeat presentation of material, requesting a slower pace, or consistently performing badly when tested etc. This tells the system that the reader isn’t engaging with the material as well as they might. If that is the case, the platform can adapt the presented material using different wording or even switch from audio to visual media. The platform then encourages the reader to-engage with the content. The benefits for an Ed Tech system here are obvious and substantial.
The system can ask questions to test the reader’s understanding of the salient points of a previous chapter, e.g. “When and how did Joseph Stalin die?”
Literal searches and 'similar meaning searches' would mean that a Search of “serial killers” would include all literal occurrences of the search (as a standard e-reader can do now), but substrings with similar meanings would also qualify, e.g. the search“serial murderers” would bring the same result.
Characteristic(s) search. Entities can be defined (in part) by their characteristics, e.g.“John is tall and handsome”. If readers search“tall and handsome”, results will include all entities containing both these characteristics but also including John.
Expanding on content. The 'reader' can ask for more information about anything, and online connectivity enables the system to bring up other relevant media related to the query, e.g."Tell me more about Joseph Stalin..."Summaries can also present sections within books by the same author or internally, in terms of a particular query, e.g. in the novel 'Emma' by Jane Austen: "When does Emma Woodhouse first meet Jane Fairfax?". Or, perhaps: "Does Jane Fairfax appear in any other Jane Austen novels?"
The client reported extremely positive results, which, due to the autonomous learning of the AI, verified by sentiment analysis and other feedback, only improved with the number of publications indexed.
A system of interactive ‘Q&A’ was piloted by one of the client’s educational institution customers, and very promising initial results were noted.
The vCAIO will conduct a situational assessment to identify the needs of, and opportunities for, the organisation. They will provide a report outlining a gap analysis and the baseline from where to start defining the AI strategy going forward
Create a vision of where the organisation should be going
Set standards for strategy implementation through establishing the correct management structures and considering ethical issues
Make things happen by operating at the right level, knowing what to outsource and what to achieve in-house for optimum delivery
Build capacity in the organisation by recruiting the right talent and identifying appropriate vendors or consultants to solve key problems
Become the public face of the new technology, both internally and externally
Understanding your needs and identifying opportunity to exploit
Evaluate your data
Knowledge Gaps | Costs v Benefits | Your Competitors
Latest Industry Tech and how we can enhance it just for you
Our key recommendations and suggested next steps