Fredrik Schill & Jens Halvarsson, Textalk: Take Part conference. Stockholm, May 2016
Hello, everyone. My name is Fredrik Schill. This is my colleague, Jens Halvarsson. We are representing Textalk. I would just first like to give you a very brief outline on the agenda for today. We will start by giving you a short introduction on who we are, where we come from. We’ll also give you a short background on some of the projects we have taken on before that has led us to draw the conclusions that we have and are presenting in this.
After that, we will continue with a product that we are currently working on, where we try to implement these ideas. If the technology is with us, we will also try to finish up with the live demo on where we are with this right now. If you would like the presentation, you are welcome to contact me later by e-mail. You can reach me on email@example.com and I will be happy to give you the presentations.
Some basic facts on us. We are representing Textalk. It is a Swedish privately-held company. We come from Gothenburg which is the second city in Sweden. We are on the other side of Sweden, on the west coast. We have today, approximately 50 employees working on the company. The reason why I mentioned that we are privately held is because it’s important. It’s makes us independent. We have the possibility to choose what projects we take on, that we do have any foreign investors implying what we should do or what conclusions to draw.
We are active in essentially two business areas at Textalk today. One of course is the media area, which this presentation is about. It might also be good to know that we are a big e-commerce partner in Sweden. We have a web-based platform for having e-commerce shops. We are today hosting approximately 4,000 web shops in Sweden, actually makes us the top supplier in Sweden in that field.
The good thing with that is that it gives us a bit of financial strength and also some wider technical knowledge that we wouldn’t have if we were only in the media business. It also gives us some connections perhaps in some business connections essentially that could be useful when we would like to combine these two areas.
The background for us goes back about 40 years. We have a long history. We are a spin-off company from a research project that was initiated in the late 1970s by the local newspaper in Gothenburg called the Gothenburg Post. It was a project initiated by them, aiming to create an all digital newspaper for visually impaired and blind people. The aim was to have something that could cover the whole content of the newspaper, all editorial content as much as possible of all ads as well.
Back in the 1970s, creating such a system was a research project. They contacted the local university in Gothenburg. They met a person who was equally interested in this, who later became the founder of our company. That is the story. This project was a success. They worked on it for a few years. It was released, the first version, in 1979. It got quite much attention. It was, as we believe, actually might have been the world’s first digital newspaper to be on a daily basis, distributed in a digital form, actually, not just for visually impaired but for any audience.
There were of course, a lot of challenges back then. Then through the 80s and 90s, we continued to spread this system for more titles in Sweden, domestically and also in approximately 10 more countries in Europe. Today, we are involved in a project that we are very happy to be involved in, which is a project run by the MTM here. It is a nationwide newspaper project. We’re running a newspaper service, which is funded by the Swedish government. It’s run by MTM and through a procurement, we are responsible for the production and distribution of it.
Today, we are covering about 120 newspaper titles in Sweden. We are reaching about 5,000 visually-impaired and blind readers on a daily basis with this. One challenge that we have seen through these years and through these projects is the challenge to get the material from the publishers. This touches quite much on what they’ve spoken about also. The key point is, of course, to get your hands on the material in a structured way, that is, of course the challenge. It has been, essentially, equally, complicated from the 70s up until, and still is today.
The aim we have had here back then, and also now in this project has been to have high-quality projects and, where you can expect to get all the editorial content and all ads and things. In that case, it’s not sufficient to rely on an XML feed that perhaps a newspaper publisher would supply. You could be lucky and that could be good. In most cases, it will be of fairly poor quality. What we decided on, we need to work with something that we can expect to have a high quality and have all of the content. We decided to use the print documents as the input source for this. We’re working with InDesign, Quark or PDF documents, whatever we can get from the publishers.
With this, we have created the production system because what those documents are lacking is the reading order of the text and also the segmentation into articles in a correct way. That’s not part of that. What we have done is we have created the interaction tool where we perform a small manual test every evening before publication of the titles. This ensures so the human intervention makes sure to get this in a high-quality manner. What we achieve with this is actually a very high-quality editorial database of the newspapers actually. In many cases, for the much better quality than the publishers had themselves originally.
This is really good. We have actually accomplished something that can get good material for us. That would probably mean that we have set up a nice system and it works. We have a good quality and everything. Everything should be good. Then the challenge is that we are a commercial company. It would of course, be nice to do this in more places than only in Sweden, naturally.
A few obvious problems exist then. One is that we have a very luxury model in Sweden here with the public funding of all the production system and all the distribution and helping even the publishers with more activity when that is needed and so on. This is something that is, if not unique, at least extremely rare to find in other places around the globe. That’s also something we saw. We had, as I said before, earlier projects in several places in Europe. Most of them were eventually canceled, not for technical reasons but for lack of funding.
Publishers also have a tendency to have a problem to focus on accessibility. If we look back to the reason why we were even started, why that project took place, that was one single extremely devoted journalist in that newspaper who fought internally to be allowed to set up that system in the 1970s. If he hadn’t done it, we would probably not have been standing here today. In many cases, if you want to bring the publishers on with this, just like that, you have to have some extremely enthusiastic engaged staff, someone at the publisher who has this incentive. That may not be so easy to find.
If you would like to do something on a larger scale, it’s difficult to rely on this to be successful. What we decided that we would try to do and what we believe to be the best solution for this, to be able to create on a large scale accessible newspaper services around the world is that we need to find a way to include accessibility into mainstream products for the publishers. We believe it will always be very difficult to create something dedicated afterward that you add on. We believe if you should be truly successful and be able to do this on a large scale, it has to be more inclusive. That’s what we think. It’s not only that.
There also has to be possibilities for the publishers to have a revenue chance on this. Publishers are under hard economical stress already. There has to be a possibility for them to see revenue possibilities, also from every subscriber. Finally, the fact is also that publishers have existing solutions today, for e-paper readers also. Many cases, they are just PDF replica readers. They are normally not accessible at all, of course. The problem is they have them. They may be in some contract with those vendors and so on. In order to be able to convince them to change into something, to adopt another system which is accessible in itself, the threshold for entering such a system has to be kept low, which means then that the fee to vendor implementing that system would have to be low so the cost for using it for the publisher has to be kept as low as well.
If it’s possible to achieve such a system, we think there is a chance to create inclusive publishing in a larger scale. That’s the background. Now, Jens will continue with the project we are working on.
[Jens] Thank you. Yes. What I will show you today is our work in progress where we’re trying to build a platform that provides publishers a great incentive to prioritize accessibility. Almost two years ago, Textalk started a new product in with a clear mission or a really hard challenge to solve some of the most common problems in the media industries. I’m talking about how to improve or simplify digital publishing, how to improve reading experience for everyone, everywhere. Of course, how to monetize digital content.
From the very beginning, we started this project by establishing partnerships with both publishers and readers to understand their needs and of course to ensure valuable output is delivered. Collaborating with our partners has been crucial to our product objectives. We have also further developed our cooperation with MTM in order to improve readability for people with dyslexia within the same mainstream solutions used by people without any reading disabilities.
This is what we came up with.
We have built something we called for a born inclusive publishing platform, Prenly. Prenly is three products in one platform. It is a publishing product, which is a content-centric management system, where all content is separated from its presentation enabling publishers to easily create, manage and repackage their content into multiple formats and probably most important, in multiple channels such as web pages, blogs, social networks, DAISY players from one place. Prenly is also an engage product, which is a fundamentally new way to create, manage and retain users, users that could be subscribers.
Users that could be one time purchasers and users that have one account for accessing all the different services these publishers provide. Our read product, I will give you a short demo of the Prenly Reader. Let’s see here. This is always scary to do a demo. I ensure you, it has worked. It’s here. This is Prenly Reader.
By using Prenly Reader, I am able to start reading from their PDF replica mode. Why? Good question, Fredrik. Because the PDF replica reflects the prioritize of content. I’m quite sure you agree it isn’t that readable or it doesn’t give me a reading experience. Since all our content is separated from its presentation, we are able to click on each article from this mode and get it in this presentation where I as a reader, are able to customize the content to fit my reading needs.
I’m able to decrease the row width which is a good feature for people with dyslexia. I’m also able to increase the font size or probably change the font size or change background color. I’m also able to start listening. Now, this is Swedish. [ 00:15:48]
While listening, I’m able to highlight word and sentences or just sentences or just words. This is just a little part of the Prenly Reader. From this mode, I’m able to continue reading within this article mode. The thing is we truly believe that this reader could be something for everybody. What I want to say is we truly believe that the Prenly Reader should be designed for everyone with or without reading disabilities.
This solution might be very good for me as dyslexia, but it might also be good for me while driving my car. I can easily connect my mobile device to my Bluetooth stereo and keep on listening to article while driving or while I’m out jogging. All this is in one mainstream application. One of our further requirements within this product is also to detect VoiceOver. If a user has VoiceOver activated, we will automatically customize the user interface to fit their reading needs.
They don’t need to see the PDF replica. They might even need to click on “Read today’s paper.” We’re able to customize their experience based on the user’s requirements or wishes. I think we made a new record. Yeah. Any questions?
[Person from the audience] Thank you. I’m just curious if you have, so to say, convinced any newspaper producers to go into this system of publishing in the future?
[Jens] Excuse me, I did not hear you.
[Person from the audience] Are you working with specific newspapers on introducing this way of publishing in the future?
Yeah, we are, yes. Today, we have 30 newspapers lined in Sweden with us.
[Person from the audience] Are you considering branching out into other languages?
Actually, right now, we’re supporting every language that it’s supporting within Android or IOS. Yes.
[Speaker from the audience] Did I understand it right that it’s a binary choice between PDF replica mode or a Word document? Because it says “Word” down in the corner. Is that Word?
[Fredrik] No, no.
[Speaker from the audience] Oh, it’s Word highlighting.
[Speaker from the audience] What format is behind your simplified view? This is an internal format. The format is what we think is not important in this visualization. Sure. The intention of this to have an inclusive system which has an inclusive user interface. We have tailor-made the format to fit that.
[Fredrik] This is an internal format. The format is what we think is not important in this visualization. Sure. The intention of this to have an inclusive system which has an inclusive user interface. We have tailor-made the format to fit that.
OK. Yeah. The positive thing is that when we create the material for this, we structure it in a way so that we can use it for this output, if it’s desired or we can choose to output it as we do in Sweden today, for example, for as DAISY 3 or DAISY 2 as well, or we could use the output that’s EPUB.
This is one example how this publishing … This is one output of the publishing platform which is intended to be good for the wider audience.
[Speaker from the audience] Thank you.
[Jens] Excuse me, I think one important thing to remember there. It’s a good question that is this a way to give the publishers a strong incentive to prioritize and invest in accessibility. They invest in a better reading experience for the big public. In the meanwhile, they are investing in structuring their content, allowing us to create DAISY publications or whatever the user needs.
We are actually, right now, this moment, having a small research project actually together with the MTM, with a test group with dyslectic people who is testing this interface and giving us feedback on how we can do this as good as possible. That’s something actually we are having active just right now.
[Person from the audience]: OK. Thank you. I mean, this looks fantastic. I guess, one of the potential barriers that I see is the requirement on the publisher’s side to get the content into the right shape to provide it to you.
Can you talk a little bit more about what’s required from the publishers? From the publisher right now, given the production tool I mentioned a little bit, is quite low so that we can handle InDesign or Quark print documents if they have them, that’s good because they provide metadata.
[Fredrik] From the publisher right now, given the production tool I mentioned a little bit, is quite low so that we can handle InDesign or Quark print documents if they have them, that’s good because they provide metadata.
We can also handle PDFs. That means we do small manual step in structuring it. Within this, we ensure the structure essentially. It’s about the reading order and just defining the content? Reading order and the segmentation into articles, essentially.
[Person from the audience]: Is it about the reading order and just defining the content?
[Fredrik] Reading order and the segmentation into articles, essentially.
We also have actually publishers running it right now in a system where they actually export article by article from their editorial systems. In that case, we can omit that phase. It depends on the publishers. Typically, publishers don’t have that good structure. That’s the problem.
[Person form the audience]: OK. Well, thank you very much, Fredrik and Jens.