Moving Towards Meaning

This article is about formatting of documents, the semantic web, tagging of content and liberating content. The goal? To give background to why we’re passionate about open and social learning here at Coracle.

I’m going to start by talking about an experiment from American philosopher John Searle known as ‘The Chinese Room.’ Searle first presented his idea in 1980 and at the core of the experiment is the question, “can a computer be intelligent?” The Chinese Room experiment sets out to prove that they can not and he does this by supposing that a computer has been programmed to receive Chinese characters and then to respond, after some computation, with a Chinese character. The machine is programmed to do this sufficiently well that it appears to a Chinese speaker that they are having a genuine conversation.

The question is, did the machine actually understand the conversation?

Now Searle imagines placing an English speaker in a room with a version of that programme. Chinese characters are passed under the door. Using the same programme logic as the computer had the English speaker would find a phrase from a phrase book and respond by pushing a Chinese character back under the door. The conversation may make sense, but the English speaker will not have understood it. Searle argued that simply following a programme does not therefore indicate understanding.

So what has this got to do with document formats and e-learning? Much more than you might think!

The key point here is meaning and the semantics of the information. When learning something, the goal is to gain a good understanding of the material and not just the ability to reproduce it word for word. As such, the right presentation of e-learning content is critical.

The semantic web

The semantic web is the goal of making documents (text, video, images, and so on) ‘understandable’ to machines. At present, humans must do a lot of work steering computers to carry out tasks such as searching for information, sorting data, fragmenting information and then recombining it to make meaningful new content. But if you describe a document and its different parts using labels that a computer can understand, you make it much easier to manipulate information to help humans. Let’s take an example. In e-learning, a physics course might be labelled under the topic ‘physics’ and the function ‘e-learning’. Individual lessons might be labelled according to their difficulty or their place in the curriculum, individual paragraphs could be classed as ‘example’, ‘essential principle of physics’ or ‘lesson contents’. Images could be ‘diagrams’, ‘illustrations’ or ‘atomic structure’. And so on. This means that new documents, new lessons and even new courses can be constructed on the fly according to what a learner needs. It helps learning become collaborative (social learning), and it provides a structure for storing and searching for documents and fragments of documents.

The right presentation

When I was working in research at university, one of the main tools we used to write professional documents was LaTeX. It’s a markup language with a particular focus on academic documents.

Writing a document in a markup language is fundamentally different from using your favourite word processor. A word file, for example, has a fixed layout and pagination: what you see is what you get. Writing with markup is different as you separate the document into two fundamental parts - the content and the information about how to display it.

The content is the text and the images and whatever else you might want, but the clever part here is that the content is also labelled with meaning. So a title is marked as being a title, an image caption as belonging to an image and a bibliography as being exactly that. The formatting is done later.

In the case of LaTeX, you can imagine a journal being able to publish an article in its own format without having to explain to the writer how to do the formatting. All the formatting across all articles in the journal will be consistent and the institutes that provide the articles can have their own style. This means that a document written in LaTeX is not the document that is read by the public. The LaTeX document plus a style file is used to generate a fixed-layout flat document, such as PDF or PostScript, which can then be published. The thing to note here is that the PDF is not the primary document: it has been rendered from the LaTeX document.

Can a document answer questions about itself?

A document itself cannot do this, but a device that interprets a document certainly can. How many words a document has is one simple example. When it comes to e-learning we want to ask more interesting questions such as ‘what part of this document relates to my mathematics course?’ or ‘how many students have problems with this section of the text?’ This is impossible to do without including information about the meaning of the text placed within the text itself. This brings us back to markup. Markup can be used not just to tell us what the different parts of a text are so it can be formatted later, but it can hold whatever information you want about the text. This could be keywords, links to related questions or materials or just an identification code so you can monitor what you have read.

Your role in learning

Fundamentally, learners are responsible for their own learning (you can take a horse to the water but you can't make it drink). But that’s not the whole story. When thinking about social learning, your role may be as a teacher, pupil, member of a working group, trainer across the internet, training provider, parent etc, and/or a combination of these. This is why e-learning should provide a dynamic social tool that can make sense of the meaning of content and how it is being used.

Different types of learners in different situations

We all run our lives differently and learn in different ways. Today we are lucky in that we have a large range of electronic devices to choose from to aid our learning. This surely means that the same content should be equally well presented on any of the numerous devices...

This is not a problem because documents can be provided in a markup format. Even if the content includes something fixed like a PDF, this can be easily be rendered from a markup file. This is great because documents, parts of documents, or even user selected subsections can be transferred from device to device with ease. Interestingly, the ebook epub format uses XHTML markup and CSS styling, so it’s very easy to present the same content in a web browser. Document security and format fixing is handled when a document is rendered from its markup. A service provider of content material should own the markup, not a fixed instance that would be rendered for a specific user.

HTML5

This brings us back to our friend HTML5, a modern extensible markup language that can be used to describe the semantics of the content. Not only does HTML5 include more semantic tags than its predecessors but it also includes a markup specification called Microdata, allowing the meaning of the content to be tagged more specifically. HTML5 provides a good foundation that can be easily ported to other devices and formats in a meaningful way.

In Conclusion

All this opens up amazing opportunities for the content we create here at Coracle. It allows users to choose their own goals and to personalise their learning. Our use of Tincan API means that individual learners can record their progress no matter what form it takes. Semantic tags allow us to mark up content so that it has meaning for the huge range of formats it might appear in, and offer a way to adapt and tailor courses in real time.

In short, we use this knowledge to help liberate course providers and individual learners from the restrictions of formats and devices and we look forward to talking with you about how we can help you.