>> KLAUS MIESENBERGER: Thank you very much. Welcome, ladies and gentlemen, to the Easy to Read on the Web symposium. On behalf of Kerstin Matausch and Andrea Petz, let me thank WC3 and WAI Research and Development Working Group for the terrific support they gave us in putting this symposium together.
I will give a short introduction to the symposium, and later on for the three parts of the symposium I will hand over to the other chairs, Kerstin and Andrea, to lead the discussion.
First, a short introduction. As we all know, globalization and universal use of the Web brings forward the need for a better understandability and readability for different user groups, language, and having different language and cultural backgrounds.
In general, this is often addressed with concepts like plain language or text customization. Text customization as most of you know is one of the symposium items that was discussed several weeks or days ago.
For specific groups like users with cognitive disabilities and sometimes also for the aging population, the concept of easy to read is used besides concepts like plain language text customization.
This symposium addresses the issue of easy to read with the goal to set the ground how to improve guidelines, techniques and tools for Web accessibility, and in particular how to reflect easy to read in the Web content accessibility guidelines.
We called for contributions related to questions as how to define and understand easy to read and what are the needs of end users.
We want to collect the approaches and concepts which are used in practice and which would outline the state-of-the-art. And we considered and we asked for contributions how to integrate easy to read into the bigger picture of Web accessibility related to reflecting it to the Web content accessibility guidelines, how to provide guidance and support of developers, content authors and other stakeholders, how to set up and how to brush up tools, helping in day-to-day practice or for specialized experts and user group domains and in mainstream; finally, what research and development efforts are needed to develop this issue further.
The call was answered by many informative and innovative contributions. And after a review, the accepted papers such as [inaudible] symposium in three parts, the first one focusing on easy to read guidelines and the impact that these guidelines to Web content accessibility guidelines.
This topic will present three papers. The second one focuses on tools for easy to read, which has accepted 7 papers, and finally a part which focuses on work flow process and services for easy to read, with four papers included.
Each domain will be chaired by one of us, a symposium chair. We will ask questions, and authors are invited to relate the papers to the key questions of the symposium.
As the audience [inaudible] advanced level of background to make the best use of the symposium. Therefore, the IRC channel is ready that you can pose questions.
Then we can later bring into discussion the [inaudible] therefore, I hope you have your coffee ready and we are ready to start.
With this, I will like to hand over to Kerstin Matausch to guide us through the first topic, easy to read guidelines and impact on Web content accessibility guidelines.
>> KERSTIN MATAUSCH: Thank you.
Thank you for the wonderful introduction. I'm glad to be able to guide you through the first section concerning guidelines.
We have a rather broad variety of guidelines to read, national guidelines, we have trans national guidelines, and we have a request to also integrate with cognitive disabilities on the Web, and regarding this we have how to consider them into the WCAG.
In this part regarding guidelines, we have three papers. And I would like to introduce very shortly our panelists. These are Annika Nietzio from FTB. She wrote the paper regarding easy to read and plain language, with the title defining criteria and refining rules, terrific work on the question whether the guidelines are compulsory or whether one could develop them further.
Next we have Leealaura Leskela from the Finnish Association on Intellectual and Developmental Disabilities, from the department, Plain Language Center. She wrote a paper on the question, whether we have only guidelines or even standards for easy to read, and whether this is sufficient or advanced. Last but not least, we have Sami Alli also from the FAIDD, Association on Intellectual and Developmental Disabilities.
He is from the department Papunet Web Service Unit. He wrote about some challenges for developing an easy to read Website, and wrote about the WCAG 2.0 for estimated, that they don't always match user needs in all situations.
So, my first question goes to Annika. In your paper, you focus on defining criteria and refining rules also with regards to Web application. What is your opinion regarding the question which aspects are necessary requirements that should be integrated to WCAG, into which in your opinion are more up to now?
Could you hear my request, Annika? Hello?
>> Sorry, I unmuted the wrong line. The question was for Annika.
>> KERSTIN MATAUSCH: Shall I repeat it?
>> Hello? I was answering all along, but maybe nobody heard.
>> KERSTIN MATAUSCH: It seems so. Could you please repeat your answer?
>> ANNIKA NIETZIO: Sure. In our paper we looked at the, really made a difference, association between easy to read and plain language. When we look into WCAG 2 as it is defined today, we see, for example, in the access criterion 315 about the reading level that the text should not require reading ability more advanced than lower secondary education.
This is something that is closely related to plain language. But looking at easy to read, this is a bit more. So we need text that is even easier to understand, even lower reading level than the lower secondary education.
For the coverage in WCAG, we can see that a lot of aspects of easy to read are already covered, for example, some aspects about language. We also have some aspects about the presentation, and but these are usually only at the level AAA or some of these are only part of the advisory techniques.
So it is there, but it is really only at the highest level of WCAG. And what we found in our paper is maybe similar to what the relationship between easy to read and plain language, you apply similar rules. But for easy to read, there are fewer exceptions.
For example, you have to apply the techniques more strictly. There are fewer exceptions, and there are .... (pause).
>> KERSTIN MATAUSCH: Okay.
>> ANNIKA NIETZIO: And then -- sorry.
>> Annika, we can't hear you.
>> ANNIKA NIETZIO: Okay, I couldn't hear your question. What were you saying?
>> KERSTIN MATAUSCH: Just proceed on, just go ahead with your answer, please.
>> ANNIKA NIETZIO: Maybe to wrap up what would be our input to WCAG 2, first of all, maybe we need a clearer definition of the scope of easy to read in WCAG 2, so for example, which information on the Website must be available in easy to read. There is already an advisory technique saying, providing text for navigational and landing pages, that requires reading ability less advanced than the lower secondary education level.
This could be a starting point for defining the scope of easy to read within a Website.
And of course, as I already said, many of the aspects of easy to read are only covered in the advisory techniques, and that could be further elaborated, for example also using the approach described in our paper.
And for the optional issue that is again about the priority, many relevant success criteria are only on level AAA. And it will be very important to take needs of persons with learning difficulties at a higher level, what level A or AA already.
Maybe as the last point, could also add more detailed techniques, not only for the text part or for the language part, but also for additional aspects that are important for people with learning difficulties; so for example techniques for the presentation, navigation and interaction that describe what has begun to make the Website really easy to use for these target groups.
>> KERSTIN MATAUSCH: Thank you very much. My next question goes to Sami. I would like to hear your opinion on that, and especially do you think there is a line of what you think it could be difficult to integrate easy to read to the WCAG?
>> SAMI ALLI: Well, hello to everyone. Hope you can hear me clearly. Yes. I think there is a line somewhere. It needs to be drawn.
It is not easy to engineer a set of guidelines to secure an accessible Website to call easy to read user groups. According to studies made, visual to text content, some easy to read user groups need a simple page structure, basic navigation, one or two levels. (muffled audio).
Very small amount of information or content or other elements. These users are critical to some easy to read user groups, so it can be argued that they should be included to WCAG 2 somehow.
But I'm not sure if it is possible to do that, and there are a couple of reasons for that. First, it isn't easy to [inaudible] precise guidelines for these general rules.
Secondly, these rules determine quite strictly the way structure and navigational structures are constructed. So I think the key issue here is, before we can discuss individual guidelines or aspects, can WCAG 2 serve all easy to read user groups.
If it should or can, then it can be discussed what alterations or modifications WCAG 2 needs to be made. If not, then it has to be discussed, for example, if it is possible or reasonable to make a distinction between plain language and easy to read, and include only one of those to WCAG 2, and maybe we also discuss what does this distinction mean.
Yes, this was my answer. Thanks.
>> KERSTIN MATAUSCH: Thank you. Time is already pressing now.
I would like to proceed on to Leealaura Leskela. You have worked six years on the practical execution of plain language and easy to read, especially from linguistic perspective.
So I would love to hear what is your opinion. Do you think easy to read on the Web will become a general necessity? Or will it be your opinion more or less stay a specialized service?
>> LEEALAURA LESKELA: Hello, can you hear me?
>> KERSTIN MATAUSCH: Yes.
>> LEEALAURA LESKELA: Hello, okay, good. Hello, everyone. Thank you for the question, and inviting us to the symposium.
I think that my answer is of course restricted to language and not to the technical issues if I understood you right.
I think of course we can make easy to read to be mainstreaming or be something that is a part of the general building up of a Website.
But I think that we need to think again what the guidelines for easy to read are, whether there can be such standards, or whether they are more like general guidelines for easy to read.
I have a suggestion, how we could do, like build up international guidelines for easy to read.
What we need from my point of view, we need one general guideline, that is loose and can flexibly be adapted to subcategories.
And these subcategories could be more precise guidelines.
I think we need three subcategories. And these are, we need a category that takes into account that we have different easy to read readers. We have persons with learning disabilities, non-native speakers.
We have persons with memory illnesses, etcetera. They need all slightly different language.
Then we need a subcategory in the guidelines that takes into account that we have different genres. In Web we have informing text, majority of the text are informing text, but we also have literature like easy to read eBooks, and we have easy to read news. And all these text genres needs to be written a bit differently.
Then of course, we need language specific guidelines, and this is what in Finland has been done already, that we understood we can't operate with the international guidelines only. We need something specific for the Finnish language.
This is, to create these subcategories is a huge national and international process and job. And this process, to do this job, we need easy to read readers, as proofreaders, and now I mean all groups, not just persons with intellectual disabilities, but also these other groups.
We need language specialists to say about and tell about the language structure. We need research about ability assessment like PSAA and PA, PIAAC. Then we need also specialists of different texts, journalists and teachers and authors to tell how different text genres can be written.
After reading the answers to the symposium, I think this is something we can do, to consider one general guideline for easy to read, which is then divided to these subcategories. This is my answer to the question, thank you.
>> KERSTIN MATAUSCH: Thank you, too.
To summarize, if I understood it right, you think it's necessary to do a channel guideline which can be flexibly adapted to various subcategories, including, for example, the different target groups and the text drawers but also the research. What you say is very very interesting.
I think we will hear more on this in the section about work flow processes and services.
My next question, Sami, to you. What do you think, how do you [inaudible] easy to read in [inaudible]
>> SAMI ALLI: This is an interesting question. If plain language is understood as a type of language which is quite near to standard language, it can quite easily be integrated to every day word processes and also to WCAG 2, by implementing easy to read material, if easy to read in this context means a more profound modifications to language and also the structure of a Web Page, to implement Web design process is much more difficult, and also integration of WCAG 2 is of course more challenging.
But I think one key issue here is that where there are different guidelines to plain language and easy to read as a language, so stated in the paper of Kerstin and Annika, there exists enough information on what this distinction of plain language and easy to read means in a Web Page, I mean what are the differences when a page is designed to plain language users, compared to when it is designed to easy to read users.
How, for example, should navigational structures be designed. I think this question needs to be researched further. Thanks.
>> KERSTIN MATAUSCH: Thank you for your answer, Sami. My next question will go to Annika.
We have this problem that many many guidelines related to various nationalities, various specific languages, also there is concepts and tools. Do you think that easy to read guidelines, concepts and tools could be transferred between different languages or even between different cultural contexts? (pause).
Annika, we can't hear you.
>> ANNIKA NIETZIO: Can you hear me now?
>> KERSTIN MATAUSCH: Yes, wonderful.
>> ANNIKA NIETZIO: Okay. Regarding the transferability of the easy to read guidelines and concepts for different languages, I think there are already a couple of projects who have shown that this is possible.
For example, the European pathways project has worked on the harmonization of easy to read guidelines for different languages. These guidelines are available for a couple of languages.
And so this is also, I think this is the high level that Leealaura Leskela was referring to earlier.
I think this is the level you can get same guideline for different languages.
But then of course the more specific the rules get, the smaller the chances are to transfer this to other languages. In our paper for example we looked at very specific rules that are built around dictionaries for a specific language or built around the words that are used in the language, and this is of course not always transferable, but the methods that were used to derive these rules, this could also be applied in other languages.
The concepts for deriving more specific rules should be transferable as well.
For the cultural context, I think that this is rather difficult, because all over the world, the daily life of people with learning difficulties still differs a lot. Easy to read is not a separate topic. It is closely connected to access to education and the possibility to learn reading and learn using the Internet.
I think it's an open question, how easy to read could be addressed in a completely different cultural context where also maybe the development of the disability rights is not, has not progressed that far.
So yeah, I think that's an open question.
>> KERSTIN MATAUSCH: Thank you, Annika. Leealaura, I would also like to hear your ideas regarding transfer of easy to read. Is that guidelines are more specific than standards, so what do you think? You are living in a multicultural country. What are your ideas regarding a cultural transferability?
>> LEEALAURA LESKELA: I'm not sure if I can say about cultural transferability of easy to read guidelines.
I would rather speak about language differences and transferability between languages. But principally I agree with you, Annika.
I think that what is easy or difficult in human languages is of course universal and can be transferred. But this universal concerns just the basic ideas of easy to read.
And what is specific in one language builds up the difficulty or easiness that must be done that the specific easy to read guidelines must be done in all known languages.
Just a few examples, well, English and Swedish which are those languages which have the greatest impact on inclusion in, in the ISLA guidelines for easy to read, they are broad languages that have a simple morphology and limited variation in syntax.
For example, Finnish language has a very rich morphology variation. I'll give you an example. For example, the word hand in Finnish can be conjugated in a thousand different ways. I can say (speaking Finnish) Etcetera. So all depending on the clause that I have. I have to make changes to the very word.
Same goes to sentence structures. We have more complex system than, for example, English. I give you one example again. The sentence, I know he is my friend, can be said in three different ways in Finnish. We can say ... (speaking Finnish).
All these mean the same. Now according to that knowledge, I have about easy to read, we have implement about easy to read, is that we know that the first sentence is the easiest. The second one is easy for people with learning disabilities. But rather difficult for persons who doesn't speak Finnish as native language.
Then the third one is most difficult for all groups. This was to show we need language specific guidelines. Inclusion Europe standards do not say anything about this, not even the Finnish version of these standards.
I know in German, there is, for example, in German language, there is this specific use of so-called.
For example, it means to question. If this is your difficult, this is something that should be taken up in the German easy to read guidelines. Or I know very little about French or Spanish. But as far as I know, there are in these languages specific verb conjugations like subjunctive or these kinds of differences. These need to be taken into consideration in these languages, and I know nothing about Japan or Thailand or languages, they are all specific languages.
So I think there are a lot of national or language specific questions that need to be taken up. Okay, this was a long answer. Sorry.
>> KERSTIN MATAUSCH: Thank you for introducing us to very rich morphology of Finnish. Thank you for that.
Summarizing your answer, it sounds to me that we have to find the lowest common denominator in order to come to a most trans national valued guideline, and this will lead us to the next section, which is about tools, and which will be moderated by Klaus, who will ask how we can integrate easy to read in tools and how we could integrate easy to read into WCAG tools.
>> KLAUS MIESENBERGER: Thank you very much, Kerstin. I hope everybody can hear me. I heard in the background that I was a bit scattered before. I hope it's better now.
As said by Kerstin, one key aspect of the uptake, also the quality and in particular also the transferability of easy to read is having good and efficient tools attend. Tools on the one side (sirens) for mainstream, Web designers, Web developers, which needs support and guidance to implement easy to read in standard Web Pages.
But also, there is a need for efficient tools for specialists, addressing the needs of different target groups like people with cognitive disabilities, that they get their services in getting access to information in easy to read language.
What is state-of-the-art regarding such tools? This is the goal of this section, of the symposium. What are the concepts which are addressed in research and development to push the issue forward?
I would like to structure our discussion along two to three questions, of course, again depending on the time which we have.
The first one is related to research and experiences of the colleagues. What are, how to connect tools or how do these tools support guidelines in WCAG 2.0, and the other way around, what are the expectations, what are the needs out of these tools, what WCAG should include that the tools could better fulfill the role.
I have said before, we have 7 speakers on our panel, 7 panelists. I will introduce them shortly before asking the first question.
Our first panelist is Vincent Vandeghinste from the Center for Computational Linguistics at the University of Leuven. His paper, linguistics and national language, depicts allowing a easier and better translation between words and pictographs. This paper relates the question of easy to read and understanding to symbol usage.
Vincent, my question, do you see symbol system or language as symbol systems or symbol languages adequately reflected in WCAG 2.0? Where would you see a need to further elaborate and extend them?
>> VINCENT VANDEGHINSTE: Can you hear me?
>> KLAUS MIESENBERGER: Fine, very good.
>> VINCENT VANDEGHINSTE: Okay. First of all, in guideline 1.1, there is explicit mention of symbols. So provide text alternatives for any nontext content so that it can be changed into other forms people need, such as symbols.
So there it is mentioned that things we research more specifically was how can we give the correct symbols, because when you have text and you convert it to symbols, there is a lot of remaining ambiguity in the words, which you have to solve in order to show correct symbols, so people who cannot read or have difficulty reading can grasp the correct meaning.
Another point where it could be more integrated into the WCAG guidelines is in principle 3, understandable. So guideline 3.1, make text content readable and understandable.
Maybe we would need some specific nonambiguous words or links to specific meanings of words when they are used.
I don't know the exact customs in easy to read guidelines. But here it would be important to make that disambiguation somehow.
This also, well, the guideline about robustness meaning that if you convert or translate to another medium like pictographs, then you would also need to make sure that it's correct and not simply guessing or something similar. Thank you.
>> KLAUS MIESENBERGER: Thank you very much. To summarize it is reflected basically and more elaborated, and some research on this would help to push the issue forward.
Let me then go on and introduce our second panelist, our second paper, presented by Biljana Drndarevic, and Horacio Saggion from the University of Fabra, and Sanja Stajner from University of Wolverhampton. It is lexical simplification strategy for enhancing text accessibility. They discuss in the paper automatic text simplification and how to apply the methodology for easy to read.
I'm not sure, I think it is Biljana who is on the call. I would like to put the question, you mention in your paper the limits of the rule-based approach. Could you elaborate where you see the limits and potential to overcome them for a better support of seeing automation of easy to read?
>> BILJANA DRNDAREVIC: Hello. The limits of rule based approach which is in our paper, manifested mainly in the fact that one can never be sure of every possible element of the language.
(audio breaking up).
Case of reporting ... no definite lists of Spanish, have derived a list, reporting verbs in our corpus and to certain extent using .... cannot guarantee that all cases are accounted for.
On the other hand, using a synonym substitution, word is substituted with a synonym from a dictionary, were limited by the dictionary itself. For example, the dictionary, Spanish, some reporting verbs on the list are synonymous to say. This approach is not complete.
Case of lexical [inaudible] best results as well as incorporating statistical methods, classified resources, large corpora, easier in the case of English than Spanish. Anything like English.
Consider anything like that.
>> Hello, I cannot hear you.
>> KLAUS MIESENBERGER: Hello, yes, thank you very much. Let me move to our third paper. It is from Alejandro Mosquera and Paloma Moreda from the University of Alicante. The title is improving readability of user-generated content in Web games using text normalization. There is a system called tino, searching and providing alternatives.
My question, what are the requirements and implications for WCAG 2.0 that such a tool would become usable? Would such text normalization demand from developers to follow guidelines and if, which ones? Alejandro, please. (pause).
Hello, Alejandro? (pause).
>> Hi, this is Shadi. I cannot locate the phone lines of Alejandro or Paloma. If either of you are on the phone, could you press 41, and the pound key?
>> KLAUS MIESENBERGER: I thought I saw them on the list when checking before the symposium. Sorry.
>> May have dropped off for whatever reason. Again, Alejandro and Paloma, if you are on the call, could you please press 41 and the pound key or the hash.
Otherwise, we might have to come back to them afterwards, if they become available on the phone.
>> KLAUS MIESENBERGER: Thank you very much, Shadi. Let me go on with the fourth paper, presented by Katarina Muhlenbock, Mats Lundalv, Sandra Derbring from Sahlgrenska University Hospital in Sweden. The title is, easy to read text characteristics across genres.
Our first question would be, you talk about children's literature; how are children's literature and easy to read represented?
>> KATARINA MUHLENBOCK: Hi.
>> KLAUS MIESENBERGER: We can hear you well.
>> KATARINA MUHLENBOCK: That is fine. We included children's literature because it also, easy to read format, where included in our corpus that we have analyzed. We found significant differences between easy to read and ordinary children's literature, at all deep language levels.
It's vocabulary, sentence structure and ideas that we have investigated with different methods. That is my answer.
>> KLAUS MIESENBERGER: Thank you very much. The second aspect, how would you see text characterization, how would it demand changes into WCAG 2.0, how to bring it into that it's reflected better in WCAG 2.0.
>> KATARINA MUHLENBOCK: Yeah, you want my opinion on that already?
>> KLAUS MIESENBERGER: Yes, please.
>> KATARINA MUHLENBOCK: Well, we have discussed it and we think it appears to be a bit premature to speak of new demanded changes to WCAG 2 in this respect at this point.
However, current research clearly indicates that genre around us is highly relevant in relation to text quality and readability.
We also think that when more developed identification tools and guidelines are available in this area, this would allow more precise recommendations in future WCAG verses.
But at that point, one could even envisage a request for genre tagging of text, in addition to language tagging in guidelines.
>> KLAUS MIESENBERGER: Thank you very much for this short and precise and very good answer for contributions.
I would like then to introduce Robert Tedesco and Licia Sbattella from Politecnico in Milano.
They presented the paper with the title, calculating text complexity during the authoring phase. They outline a tool called Sparta tool, an authoring support tool which provides numeric estimations of the complexity of text and warnings or advises about phrase structure, if uses were elaborated, and used as a sound approach to integrate it into different authoring experiences.
If I could ask to Roberto, Robert Tedesco, how would your tool ask for changes or adaptations of WCAG 2.0? What are the implications of your research for the development of guidelines? Roberto, please.
>> ROBERTO TEDESCO: Yeah, we could for example add to WCAG 2 a new requirement about text accessibility which requires evaluation of text complexity.
For example, WCAG could require authors to specifically evaluate the complexity of the text, and insert a numeric evaluation according to sound formula.
Of course, the name of the formula should be published as well. From our point of view we could add two values, one for lexical complexity and the other one for syntactic complexity.
Could add such values in the text itself, or in a special tag for example. Specifically I think in certain text complexity in a tag could be quite useful because it meets an automatic selection of the text.
>> KLAUS MIESENBERGER: Thank you very much, Roberto.
Let me come to the next paper, presented by Vasile Topac and Vasile Stoicu-Tivadar from Politehnica University of Timisoara. The paper has the title, evaluation of terminology labeling impact over readability.
My first question is, is Vasile on the call?
>> Not as far as I know.
>> KLAUS MIESENBERGER: Probably not. My first question would have been, could you outline which NLP, natural language processing tools have been applied? And before the meeting, he sent me back an answer which I will now present on ICTU. It outlines the tools which they used, was text for all, a terminology interpreter which is available at the Web Page, at the link outlined.
And if people are interested, they can look up for this tool. (overlapping speakers).
>> Sorry to disturb. Please read it out loud for the captions as well.
>> KLAUS MIESENBERGER: Www.textforall -- text, 4 as a number, all. Text4all.net/interpreter.GSP.
This is okay.
>> SHADI ABOU-ZAHRA: Great. Go ahead.
>> KLAUS MIESENBERGER: Thank you very much. This tool adapts Web Pages by recognizing and labeling the terminology.
The 7th paper we got for the tools section is from Thea van der Geest, Jana Becker from the University of Twente, from the Netherlands. They present MIA, My Internet Assistant for successfully reading and using Web content.
But first, a question to Thea: What is the impact on guidelines and what is the expectation from guidelines like WCAG, when taking text layout and introductions into account, how to bring such tools and such expectations into WCAG. Thea?
>> THEA VAN DER GEEST: Hi, this is Thea. I hope you can hear me well.
>> KLAUS MIESENBERGER: We can hear you well. Thank you.
>> THEA VAN DER GEEST: Okay. Good. MIA is from Africa. It is a plug-in you can use in any layover, you can say action oriented part of the Website, for example, form filling tasks. Jana has study with low literacy users in form filling tasks. MIA gives all spoken instructions. Those same spoken instructions are written out in a text balloon.
She points to the successive steps, for example, the successive fields that have to be filled in a form. She guides people through a Web task and the easy to read part of MI A is that she, the oral instructions are written in easy to read, according to easy to read guidelines for Dutch text in this case.
So you are asking what is the connection with any of, possible connection with WCAG guidelines. I would say that with MIA, as compared to others, it is not just looking at the text, but actually looking at the function that needs to be fulfilled there.
Forms are of course important for services or transactions going on, on the Internet.
I think we should go further than just thinking about an easier representation then of the text, easier formulation or easier construction of the text. But we should also think about the tasks that need to be fulfilled, that go beyond genre even.
I could imagine that we could come up with a directory of essential use cases of many very common tasks on the Internet, like the task of identifying yourself or selecting a product from a catalog, or purchasing or ordering a product.
These could be like essential use cases for tasks. You could imagine that WCAG guidelines would go further than just like changing the representation.
That is my answer to your question, Klaus.
>> KLAUS MIESENBERGER: Thank you very much for your introduction to MIA, and answering this question with relation to WCAG.
This finishes our first round related to the question of how tools relate to WCAG to guidelines in general, and what expectations they see out of the tool development and research in the field towards the further development of the guidelines.
I would like to come back and first ask Vincent, when you reflect on the state-of-the-art of available tools for symbol systems and symbol language use, how well elaborated would you value them to be applied at the larger scale for mainstream Web accessibility or specialized services for people with cognitive disabilities. Vincent, please.
>> VINCENT VANDEGHINSTE: The pictograph or symbol languages that we currently use, they are mainly containing words for daily usage.
So I don't think they are suitable for, to convert any Website, with any content.
But then they should be suitable, they are aimed at people with cognitive disabilities. So that already limits the Website which you should target. (Beep).
Yeah, that is my answer.
>> KLAUS MIESENBERGER: Thank you very much. If I could call Biljana in and ask the same question, how lexical, how do you see lexical simplification to become an efficient tool or what is the state-of-the-art of text simplification tools for easy to read? Could the approach be extended to support Web accessibility in general in an efficient manner?
>> BILJANA DRNDAREVIC: Lexical simplification there is one aspect ....
(audio breaking up).
Yield best results in combination .... syntactic ... nevertheless, have noticed, analyzing .... little changes applied consistently ... elevated rate. Have a more significant impact on text [inaudible] lexical changes would have to be combined with other ... include elimination, content, introduction of definitions of difficult concepts and treatment of ...
(audio breaking up).
On subset of a lexical module, subset of a larger apparatus .... able to talk about, possibility. Definitely I believe it's the direction.
>> KLAUS MIESENBERGER: We have some problems in getting your answer. I'm checking at the moment the scribe, if the information is there. We had some problems. If you could have a look at the scribe and outline if there are some missing items in your answer, and could elaborate on them, please.
>> BILJANA DRNDAREVIC: I was saying that lexical simplification would work best on, when combined to other text simplification strategies like syntactic simplification, eliminating superfluous content, definitions, difficult concept. What we do in this paper is (background noise) define subsets of lexical modules and lexical module itself is, other modules in a larger simplification system. Lexical simplification on its own I don't think it would be, have high enough impact but definitely combined with other approaches, simplification could result ...
>> KLAUS MIESENBERGER: Thank you very much for clarifying this. It is quite clear that lexical simplification should be receiving an aspect included into, how to say tools supporting target group.
If I could go on, and ask Alejandro if they are around, most probably not, could I ask Shadi if Alejandro or Paloma raised their hand, and if they are in the telephone conference in the symposium now.
>> SHADI ABOU-ZAHRA: No, I haven't been able to locate them, Klaus.
>> KLAUS MIESENBERGER: Thank you, Shadi.
Then I would like to put my next question to Katarina Muhlenbock, and would like to ask if text characterization would also help to address different user groups, will make easy to read text, make it easier to address, easy to read to more user groups.
>> KATARINA MUHLENBOCK: Yes, in fact, I believe that text characterization based on other feature values can, superficial measures can be benefit for specific target group.
But it depends on the general reading level. In addition to text characterization from the genre perspective, also the main related properties play an important role in providing ETR texts.
So these purposes are possible to capture with NLP solutions as well, and in fact, there are a lot of data mining techniques out there, that could be used.
>> KLAUS MIESENBERGER: Thank you very much.
>> KATARINA MUHLENBOCK: Thank you.
>> KLAUS MIESENBERGER: Let me go on then with, and put a question, related question to Roberto. In view of Sparta 2, the Sparta 2 tool, do research results show that numeric complexity measurement is a valuable indicator, if a text is close to plain language, or it is, it comes closer to easy to read. And in addition, have tests been done with end user groups? Are these target groups part of research considerations in regarding text complexity measurements?
>> ROBERTO TEDESCO: Yes, an American measurement of course does an approximation of the tool complexity of the text. Yet I think it is useful because it can be used as a simple threshold.
The problem is that current formulas are probably too simplistic. For that reason, we would need information provided by NLP tools.
We conducted some experiments with children and young persons, with Down syndrome, dyslexia and sight problems.
We proposed them newspapers articles targeted with indexes. Users that perceived the possibility to know the reliability of the text because such information submitted them to choose text with the preferred complexity. We also found improvements in readability, sorry, in reading speed and comprehension level.
Anyway, such results are still preliminary. And for sure, further work is needed.
>> KLAUS MIESENBERGER: Thank you very much, Roberto. The next paper of Vasile Topac evaluation of terminology labeling, impact of readability, unfortunately, he did not give me an answer via E-mail to this question.
Therefore, I would like to go on with the MIA system, and would like to ask Thea, how would you address users with different language skills? How are they addressed in such other based tools, those using different levels addressed like plain language and easy to read?
How would you see the way forward?
>> THEA VAN DER GEEST: I think that MIA already uses different systems to address people.
For example, originally written specs is more created as a dialogue, with steps and turns.
And the text balloon that Mia has, with spoken text, and it is written at the same time. So people can read or can listen to the spoken text and read the same written text at the same time.
So it has different, like different text modalities to work from. What we were amazed by when we tested this is the level of involvement that people really have with MIA. It surprised me myself. And it sounds like a bit over the top, but people really thought she was a person. They worried about her shoes, that those shoes were not allowing her to stand as long as she needed to explain the form.
She had high heels on the avatar. So people really felt like they were interacting with a person, that personal assistant and that helps.
>> KLAUS MIESENBERGER: Thank you very much for this very clear and, clear example.
This finishes our second round of questions.
I would like now to open the third round, and would like to go to a question which has a little bit already been raised in our first round about the guidelines.
I would like to ask the panelists about how tools support or integrate into the considerations of transferability into other languages, other cultural contexts, and application scenarios like, for example, legal context and medical context, technical context of information.
If I could start again with Vincent, and if I could ask you, Vincent, do you see symbol systems and symbol languages, first of all, more as a tool for a specialized translation service for a specific target group?
Or do you see it also as a tool which could perhaps not now but later be used for mainstream Web accessibility?
>> VINCENT VANDEGHINSTE: Well, I can say that we are planning a new project where we are targeting more mainstream Web accessibility, with automatic conversion to pictographs.
So we surely target more mainstream accessibility. For specific domains, like medical or so, you might need more specific symbols or more specific symbols. What we currently use, we would like to target social Websites and things like that.
>> KLAUS MIESENBERGER: Thank you very much, Vincent.
Vincent, if I could add to that, what about the question of transferability? How do you see the transferability of pictographs between, because by nature one could expect that pictographs are easily transferred from one cultural context language context to another.
How do you see the transferability?
>> VINCENT VANDEGHINSTE: Well, first of all, the pictographic sets we are currently using in our research are already used and are developed bottom/up by the people who actually need them in traffic centers or day care, something like that. So they are transferable in the sense that some pictographic sets are used in different languages, program the beta set is used in Polish and Dutch.
But defining one symbol set for all symbol users would be like defining one language that we all should speak. So thank you.
>> KLAUS MIESENBERGER: Excellent, Vincent.
Let me come to Biljana and your work and research related to lexical simplification.
Is this work, is lexical simplification across different languages, are there experiences? Is there research and development in this domain?
>> BILJANA DRNDAREVIC: Yes, the approach is a broad-based approach, based on manually. Behavior of actual language elements, natural context. This context is different, but if the language is for example, very productive pattern in Spanish ...
(audio breaking up).
Verb which would be something like [inaudible] this pattern does not exist in English, the prepositional phrase without the verb, according to the subject. Some patterns could possibly match but would also have to be rewritten for a new language.
There are other approaches, such as [inaudible] these may be more easily ported from one language to another, where the only thing that changes are the resource such as a dictionary. But obviously here would depend on the systems as such. Resource languages.
(audio breaking up).
>> KLAUS MIESENBERGER: Once more, could I ask you to have a short look at the scribe, because we have some problems in getting all the words you were speaking, and perhaps if you see where the audio was breaking up, to add to that if you see that there are considerable gaps in the answer.
>> SHADI ABOU-ZAHRA: If I may intervene, your sound is very low; we can barely hear you. Maybe if you can speak up, maybe get closer to the microphone, and shout a little more. Thank you.
>> BILJANA DRNDAREVIC: Okay. I'm sorry about that. I normally have a problem with that. I was saying that the rule-based transformations used in our paper are derived manually. By observing the behavior of actual language elements in actual context, this context is different for languages obviously. I was mentioning an example of a Spanish pattern, which would be something according to a verb which does not exist in English and English uses a prepositional phrase without the verb, plus subject.
Some patterns could be matched, between languages, but not all. The rules would have to be rewritten for another language.
Other approaches to lexical simplification such as simplification based on synonym substitution may be more easily ported from one language to another, with the only thing changing being the resource such as a dictionary. That would depend on the existence of the dictionary and the quality of the dictionary and resource in general in different languages.
It is not an easy task at all. Thank you.
>> KLAUS MIESENBERGER: Thank you very much. This is quite clear now, that the rules have to be defined, but the resources and words, that there could be a transferability of exchange of resources. Thanks a lot.
I would like to put the, according question, to Alejandro and Paloma. They are not available unfortunately. This brings me to Katarina Muhlenbock. And I would like to ask you if text characterization, is text characterization transferable across languages and cultural context? And could this be a resource, a potential for cooperation and pushing easy to read forward?
>> KATARINA MUHLENBOCK: I think we have to make a distinction between methods and resources, because our theories behind our approach, they are based on research performed as human tests on English, Swedish, Italian speaking persons.
We think they might be, or we know they might be a view of language processing in general because they say something about cognitive processes. But one NLP implementation, natural language processing implementation which is described in our paper, it strongly is dependent on appropriate language specific resources.
So in order to transfer our science to another language, you have to adapt specific corpora, word lists, lexica, word or specific tags for that specific language. That is our point. It is extremely important that you have got enough language resources in order to make something clever and efficient for other languages as well.
>> KLAUS MIESENBERGER: Thank you very much. I think this adds very well to, on the one side, the possibility to use and to have available a corpora information lexica and so on in a certain language, but in terms of the transferability rules and further issues have to be addressed in specific languages.
Roberto, Robert Tedesco, could I ask you, is text complexity measurement where tools you are working on and to your knowledge the research in this field, is there transferability across languages and cultural contexts? And in particular, what in that context are the challenges?
>> ROBERTO TEDESCO: Yeah, our indexes can be ported to other languages. In particular, if you can find a shallow [inaudible] lexica resource containing a list of words for a particular language, then our indexes can be ported to that language.
Maybe calculating the weights we use in our indexes is the hardest challenge, because our indexes basically are composed by weighted means.
So you need to adjust those weights in order to adjust the values of the indexes. By the way, even in Italian we still are working on that problem.
>> KLAUS MIESENBERGER: Thank you very much, Roberto, for this answer.
The next contribution which I got in written format from Vasile Topac, the question was, natural language processing tools by nature outlined for different language contexts, are they transferable? And once more, let me put his answer into the IRC channel and read it to you now.
He answers that the tool presented here like many other tools with terminology do transport to other domains, having direct dependences on the glossaries and training data of corpora. However, tools that are rule based and care about the grammar and language are usually language dependent, also based on the nature, the scope of the text, informative technology, use oriented, people, medical recommendations for example and others, and with the risk associated with reducing text authority or text readability we can recommend more language processing tools that replace terminology or only label it.
This is the answer which I got back from Vasile Topac on the question related to transferability which seems to be very much in line with what the other speakers, panelists presented.
Thea, if I could ask you in terms of transferring tools like MIA to other contexts, and also to other application scenarios in rule planning in particular for standards tasks, like getting, making information on the Web accessible. How do you see this question of transferability?
>> THEA VAN DER GEEST: The avatar MIA is certainly transferable to any other form. It is now being tested in the context of people applying for unemployment benefits at the Dutch state.
I think it's more designed for increasing the accessibility of procedural tasks like filling out forms or performing transactions.
Straightforward, long sections of text, reading long sections of text, because it's not only [inaudible] more easy to read and easy to understand text version, it is also pointing to the different places where people should act on or interact with the computer program.
That is why I think it's more geared towards license forms.
>> KLAUS MIESENBERGER: Thank you very much, Thea. This brings me to the end of the second part of the symposium, which focused on tools for easy to read.
I think we got and we were very pleased that we got so many feedback to our call asking for an overview and for research which is going on in terms of developing and providing tools for easy to read and readability and supporting authors, be it mainstream or also be it authors working for specialized services, developing information, or translating information for people with cognitive disabilities or other groups benefiting considerably for easy to read on the Web.
We also have outlined and gotten an overview that other aspects like pictographs and other tools should be integrated in somewhat, what might be called in the future a user experience and user centric approach to making information on the Web easier to understand and easier to read for different target groups.
With this, let me thank, say thank you, once more, to the panelists for their excellent work they presented, the papers they presented, and also for their invaluable input that they gave in the discussion, which helps us of course a lot in pushing the work inside the research and development group forward.
Thanks again once more. Let me hand over for the third part of our discussion to my colleague, Andrea Petz. Andrea, please.
>> ANDREA PETZ: Hello. I hope you can hear me (Beep) clearly. There were some issues with my connection.
My name is Andrea Petz. I have the pleasure to chair this third part of our on-line symposium. I work at University of Linz for the institute and I'm responsible for the service provision for students with disabilities. Besides this my research and teaching focus is on accessibility and usable information systems and documents, comprising easy to read and designed for all.
The overall topic of this last section is work flow, process and services for easy to read, bringing this symposium (Beep) necessary underlying guidelines over tools supporting the implementation, through possible practice.
The overall aim of this section is to give analysis of the state-of-the-art of readability and legibility on the Web, describing ongoing research activities and practical examples, and identify lack of research on easy to read on the Web and lastly, presenting concepts and models of easy to read in practice.
With me are waiting four panelists: Michael Schaten, Klaus Miesenberger, Timo Overmark and Clayton Lewis. Michael Schaten, from TU Dortmund, provided a contribution on accessibility 2.0 dealing with social Web application that was if better designed for people with cognitive disabilities already in the very beginning of the software development process.
Additionally, a community built up during this activity will compose explanations or easy to understand alternatives on complex words, terms and phrases, to overcome traditional barriers for this target group using Web 2.0 applications to a glossary of service.
The next panelist will be Klaus Miesenberger; he decided that he has not talked enough today (chuckles) from University of Linz in Austria. He presented a paper on including easy to read legibility and readability into Web engineering, where he discussed together with his coauthors the feasibility and the possible structure to include plain language or easy to read into the process and work flow of Web engineering, combined, combining the inherent work flow used for plain language and easy to read, through the meta work flow of authoring and designing for the Web in general.
Coming back to social Web applications, our third panelist, Timo Overmark from the Finnish Association on Intellectual and Developmental Disabilities, or short, FAIDD, in Finland, presenting his contribution, social networking service for people with cognitive or speech and language impairments, finding concerning mainstream social networking services, their obstacles and possibilities to overcome them.
This built up on a project from Finnish FAIDD and Papunet Network Service Unit, and it has the goal to create an easy to use social networking service for people with IDD, that is intellectual and developmental disabilities, as well as SLI, this is the specific language impairment.
Our last panelist is Clayton Lewis, from University of Colorado, USA.
He aimed in his contribution, reading adaptations for people with cognitive disabilities: Opportunities, together with his coauthor, to bring to our symposium a number of opportunities for research that may lead to ways to adapt textural content to make it easier to read for the target group.
So you see how this last section can bring this symposium to a round figure, before we start the open discussion.
Then let's champion, let's go on. The first question for our panel is, what sets of guidelines are applied predominantly on the Web? Easy to read, plain language, or others? This should focus on different nuances of simplification levels. Perhaps following national regulations, or research outcomes.
To start with a neighboring country, Michael, can you briefly describe your view on the situation in Germany? Michael? Someone out there? Okay, Shadi, is he on the list?
>> SHADI ABOU-ZAHRA: Yes. Michael, your line might be muted locally.
>> SHADI ABOU-ZAHRA: Maybe we can try to come back later, Andrea. We can go on with something else?
>> ANDREA PETZ: No problem. I will change to Klaus. Where do you see the risks, better the challenges in Internet in plain language or easy to read in the general Web accessibility and use ability requirements? How would you address these risks or challenges, and where would you draw or would you draw the line through specialized services, to specialized services?
>> KLAUS MIESENBERGER: Thank you very much, Andrea. Can you hear me?
>> ANDREA PETZ: Yes, of course, thank you.
>> KLAUS MIESENBERGER: Very good. The challenge in my opinion is first of all when we talk about plain language to take up, there is and I think we all experienced the same with Web accessibility, there is always the risk although we might have awareness, although there is policy, legislation, guidelines, techniques, tools might be there, but take-up and implementing, implementation in mainstream is not there.
So the question is, how we can foster and support that really the issue is taken up, be it on the one hand for mainstream implementation.
And I think there is a strong need that we include such issues like easy to read or plain language into the work flow of Web developers having a close look on the work flow, on the development flow of Web information provision and to integrate into this, it has to be part of the day-to-day experience to follow that guidelines, that they are implemented.
And in terms of providing specialized services for targeted user groups, I think for both aspects, both from our, both for plain language and easy to read, we need tools which help the Web developers, the content providers, that they can easily implement, be it easy to read or plain language.
And in terms of the questions drawing a line between them, to my experience, there is for sure a need and it might be difficult to implement or to demand or ask mainstream to ask for such an individualized level of easy to read.
And what I learned out of practice in analyzing the approaches and practice, easy to read is very much individualized and focused on specific user groups.
And therefore, I think yes, there is a line. We need something like a general set which we include in WCAG, and there is a specialized set which is used to address very specific user groups.
>> ANDREA PETZ: Okay. Thank you, Klaus. So if I understood it right, the work flow to design easy to read or plain language information should be, come of some kind of guideline or standard?
>> KLAUS MIESENBERGER: It could be formulated in different way. Yes, it could be part of something like guidelines or recommendations, or it could be even integrated into work flow support tools.
There are, for software engineers and also Web engineering, there are tools out there providing support to follow a certain step by step procedure in the development process. And easy to read, of course, should enter into this work flow that this will take into account, or at least be considered in practice in mainstream.
>> ANDREA PETZ: Thank you, Klaus. Nordic countries have reputation to be ahead not only in this area. Can you describe the situation in Finland and tell us how improvement may risk and within this presented project might relate and impact on existing standards, such as WCAG 2.0?
>> Hello, can you hear me?
>> ANDREA PETZ: Yes.
>> TIMO OVERMARK: There is no legislation what authorities provide easy to read content in Finland. Although the legislation on equality could be interpreted that way, there isn't such which explicitly considers easy to read.
But despite that, some authorities have easy to read sections on their Website although. What comes to be WCAG standards the service we created is mainly in that people with intellectual disabilities, so WCAG provides more of a good starting point and as accessibility for people with cognitive disabilities more or less equals use ability, we mainly put a lot of effort making the service easier to use and the structure easier to understand, focusing on, for example, the visual layout and the structure of the service and the understandability of the service structure or the purpose of the service as a whole.
And also, we try to simplify individual complex functions, such as the log-in, and the uploading of files and things like that. We broke up complex processes in smaller parts, or provided extra information, help text to assist people in performing those tasks.
>> ANDREA PETZ: Thank you. I heard Michael could be back again. But before handing over to Clayton, I will ask again Michael, can you briefly describe your view on this situation, in Germany?
>> MICHAEL SCHATEN: Sure, hope you can hear me now. Right?
>> ANDREA PETZ: Yes!
>> MICHAEL SCHATEN: Sorry for interrupting. I accidentally pressed the wrong button and muted myself. (chuckles).
I hope it will not happen again.
My opinion on the situation of easy to read in Germany, we have of course on the one hand that everybody wants to write and provide information in accessible way.
So the language, methods, message you provide is understandable to the readers, and if you are a company addressing customers or if you are a governmental institution and talking with citizens.
And we actually have several information sources for guidelines, informing us about easy to read language, how to write easy to read language especially the network of several stakeholders actively spreading knowledge about easy to read language and the whole German speaking area.
Especially, the German branch of network people is very active in this area in Germany.
My opinion, the most popular and best toolable guideline consists of two books, the one hand, books of rules of easy to read language. It has first of all a short introduction about what easy to read language is actually and how to write text in this way and so on.
In addition, it is worthy for people who get in touch the first time with easy to read language, it provides several practical examples which are very useful for a very wide, variety of applications.
On the other hand, you have a real vocabulary book for translating difficult to understand words into an easy to read alternative. Okay. It's quite a few years old but still it is very helpful in my opinion.
In addition, we have several information sources like, for example, the WCAG 2.0, provided a link to Europe inclusion in the chat a minute ago or an hour ago now, but although we have all these guidelines and information about the need for easy to read language, the most people still use quite complex wording in their texts.
I have mainly two reasons for this. On the one hand, sometimes the issues we are talking about need to complex vocabulary to let the transmitted information be precise, especially in science and law discussions this is very common.
On the other hand, I think many people just like to use language as some kind of style element. I guess they don't like the limited features you have when writing or talking in easy to read language. Especially when it comes to the one information percent it is very hard to fulfill this rule, and it regularly results in short and very staccato sentences.
So, in my opinion, easy to read language is still unfortunately some kind of an exotic feature, potentially when it comes to Websites and Web information, hence we can only find it almost on special Websites or mainly two special Websites.
On the one hand, special interests sites, so for example the previously mentioned network for spreading easy to read language, so these are sites written by or for people with cognitive and intellectual disabilities, and of course that is on the site and on the other hand we have some governmental sites which are forced to offer, easy to read language information at least for a small portion of the Website due to we, that is the regulation for accessible information technology in Germany.
But even these sites offer easy to read language as some kind of special alternative to their standard writing style, although and I suggest this is the most countries the same problem, the standard language used on governmental sites is everything but easy to read for people without any disabilities influencing the text comprehension.
Our research activities, I though hope that easy to read language will gain popularity and will find more application. And I guess symposiums, activities like our symposium today offer good benefit on this. That is my answer.
>> ANDREA PETZ: Okay. Thank you. Thank you for this overview, Michael.
Now to you, Clayton. In your paper, you cited that the best interaction is the one you don't have to have. I do hope that the interaction within this symposium is not too bad for you. (chuckles).
Can you briefly describe the situation in the U.S., that also have a long history in facilitating diverse needs of user groups. And where do you see an immediate chance for the findings you listed in your paper.
>> CLAYTON LEWIS: Can you hear me?
>> ANDREA PETZ: Yes, well.
>> CLAYTON LEWIS: In the situation in the U.S. it is not so different in substance to what was just described for some of the countries in Europe.
At the moment, there are no generally accepted regulations of Web content in the Web at large, that is in the commercial Web.
But this situation may change through ongoing legal action.
But today, for commercial applications, everyone is free to do what they want to do. For the government the situation is different in principle.
So there are plain language regulations that are in force for the federal government, and for some other units of government as well.
However, it's questionable how much impact these regulations are actually having. So agencies are required to set up some kind of policy and identify a contact person. So I just had a look at one agency, and indeed, they had somebody who is their sort of plain language person, and they invited members of the public to comment.
And they got a bunch of comments. And there was no indication that I could see of any response to these comments. It is a kind of a paper following of rules, I would say not accompanied by much impact on what is actually happening in Websites.
Again, there are occasional exceptions but pretty rare as people in Europe have been saying as well.
>> ANDREA PETZ: Thank you. This brings me to the next question to our panel.
This next question deals with an often-raised issue. We already had it today. I think we can be really brief on this, not to be too late in time.
So what about the transfer of easy information in different languages, and the transfer of work flows and already existent structures to different cultural that leads to legislation and so on. I want to check your opinion, knowing that it was discussed already.
So perhaps we can also add to the knowledge we have already, let's start again with Clayton, as you are used to talk now. Do you see the necessity and a chance for the transfer of easy to read materials to other cultural context like subject area use cases, and if yes, what are from your point of view the necessary prerequisites for it?
>> CLAYTON LEWIS: Yeah. I think that this has been covered pretty well in earlier comments. There are plainly plenty of problems, plenty of reasons why many of the techniques involved are quite specific, both content areas and to language.
It seems to me nevertheless there may be the prospect of some transfer at a higher level. So I hope we can discuss later some changes in the way people might think about this whole area, that perhaps would have relevance across different countries.
For example, the material Michael was just mentioning about easy to read, my guess would be that at a high level a good deal of that material would be shareable. But getting more specific, I think the prospects are not very good.
>> ANDREA PETZ: Yes. Wonderful. Thank you for this really brief and concise answer.
Klaus, following your research and experience, is it then reasonable to consolidate already existing principles, guidelines and work flows to one international standard for all or do you see obstacles in here?
>> KLAUS MIESENBERGER: I think there is, as it was mentioned before, there are several guidelines ought there, and several rules which are under discussion, and they have the potential which could become a standard which could be aligned at international level.
But as a second level I see what also might be of interest to come up with something like a standard in the ISO 9,000 series that we define it more on a process which helps to come to group quality of Web easy to read, so that we define the process which includes the writing, which includes the different checking of grammar, of the vocabulary and so on, and that this could be included in something like a process standard, and what would help be it at the plain language level, but also at the level of easy to read and orienting at very specific target groups, would help to reach the goal of better understandability, both for mainstream again, but also for services providing translation services for target groups.
>> ANDREA PETZ: Okay, thank you. Going from the general to the more specific, Timo, how did you develop the plain language for the term and references in your network service? And what obstacles did you observe and how did you deal with complex structures and processes? For example, log-in or [inaudible] perhaps this could be some kind of use case.
>> TIMO OVERMARK: First of all, in the project we had, we had a design partner from a day center, a group of young people with IDD.
We cooperated with them in the design phase. We started up with a prototype service, based on open source social networking software, and we tested and improved it in cooperation with them.
We tested all functionality within the site with them. We usually created a prototype version and then we asked them if it was okay or if there was something to improve and then improved it accordingly.
The user interface was developed in cooperation with user interface designer and a plain language professional. Then we tested and reviewed the language also as a part of this service, with the cooperating design team.
We also have plain language terms of service. And that was also developed and tested with the end users. The terms of service are mainly like sort of a, what is the right thing to do in the service and what is not.
We illustrated those with a comic strip, sort of like social stories. And the social stories were written with the end user design team, and the illustrations was also made by a person from the end user, user group.
>> ANDREA PETZ: Okay. Thank you. So you got feedback from the user group, the whole design group, if I'm right now, if I understand.
>> TIMO OVERMARK: Yes, exactly.
>> ANDREA PETZ: It is following rules or something like this already. To conclude this last topic section of our on-line symposium, we could shed some light on easy to read and its underlying work flows.
Michael, from your perspective, is there a chance to transfer the work flow to implement it to design your service, as well as to find and solve complex pieces of textual information to an international standard or some kind of process, other people or software designers could follow?
>> MICHAEL SCHATEN: Yes and no, in the sense of my concrete project, the question can be discussed from two points of views I guess.
The first end, the software design and development phase, so developing the software itself, and the second step, on the other hand, the operation of the developed software. Regarding the development, my project aims at a user centered design that is key functionality or key impact on it. This implies to interact with the potential future users as early as possible.
In my concrete project I did some research about when to first interact with my target group of people with intellectual and cognitive disabilities. In the end I decided to do the first exchange with them after finishing the first prototype of the software.
In this way the users have a clearer idea about the later product, and a likely, are likely not confused by some abstract idea I could only produce to them theoretically. With this approach they have an actual product and have input into what elements they would like changed in future versions.
I'm very sure this is the only sufficient way you can design and provide decent software products, show enough examples what turns out when developers do not closely collaborate with customers.
From my project, a group of people with intellectual and cognitive disabilities, special requirements result. It is useful to interact with them during your development activities already.
Yes, I guess several classic development procedures and concepts should be reconsidered to form new standards. And in my opinion, rapid prototyping and evolutionary models are well suitable models for the every day work in software development.
On the other hand, we have the evaluation (Beep) operating software, and there are several questions that I need to consider for my concrete project. For example, how usable is a service for my (Beep) group. Do the users understand the work flow I implemented and thought of?
Implemented security functions for avoiding abuse of the service are a great burden for my clientele. I expect most of the functions I used with my project are present in other software systems or were project too, especially authentication mechanisms.
I'm surely going to publish the experiences I will make during the operation of the software. But I'm not sure if this will be any more than experience report. In the end I do not expect my experiences will be general enough to be applied to any software project in a kind of standard.
First of all, the development of the software, yes, that could of course be generalized to some kind of standard. Although there are already some standards.
But for evaluating the operational software, I guess, my project is way too small to be honest.
>> ANDREA PETZ: Okay. So you play a little bit the -- okay, no. Timo, can you say really brief, as we are running out of time, what from your point of view do developers of social networking platforms need to do to better address accessibility and usability of their services?
>> TIMO OVERMARK: Well, most importantly, I think, does it focus on use ability and improving usability of the service. This will benefit everyone, not just easy to read user groups.
The structure of the service, for example, must be thought out to be as clear and logical as possible. For example, is it obvious what each individual function does, and also privacy because it's important specifically for people with IDD who shouldn't show content, content in a service, all around for service unless it's made very clear to the user who will see it.
And also from the user perspective, who is the intended audience of the things they write, who should they write to, and what to say.
And also, very important, is to verify the accessibility of the service with the user group, with the end users, by doing use ability testing with this user group, and maybe preferably have the users already in the design phase.
>> ANDREA PETZ: Thank you so much, Timo. These are known fact, perhaps over a lot of symposia, we can bring it into mainstream, and perhaps also [inaudible]
Website and information design, be able to do some kind of mainstreaming here.
Last up, please, Clayton, you provided a list of techniques to make information and services accessible and usable (Beep).
These also suggest analyze work flows, ask for a very brief, really brief intent of this roadmap for this area. Whom would you involve, what would you do next? What are the possible milestones? And where would you involve standards and guidelines, for example, WCAG 2.0?
>> CLAYTON LEWIS: Well, time doesn't permit really getting into this, but just some high level notes. One is, there is a more or less traditional agenda that one could elaborate.
And particularly in a long-term meeting when we have a chance to really make fundamental progress in models of language comprehension and so on, this conservative agenda might make sense.
But there is another one that is more radical that responds to what Michael and others have been saying, and that is that the real challenge I think is for organizations to think differently about their communication function.
Much of our technical focus is as if there is something someone can do when they create content and when they have done that, they are finished and our job is to just get them to do that right thing.
But I think we can all recognize that that is really a mistake. People have to look at the entire communication process. It is not finished when some content is created. It is only finished when communication actually occurs.
Organizations need to rethink their communication processes to include actual engagement with the people they are trying to communicate with, not least to determine whether that communication has actually happened.
This includes things like testing. But really it goes beyond that to a more important rethinking of the communication function.
>> ANDREA PETZ: Okay. Thank you, Clayton, and thank you to all, the other panelists for sharing their thoughts and findings.
A big hand also to you, the audience, for your interest in this important topic, and for your patience.
We will have now a short break to digest what we heard, and after this break, we will come back, and you will have the opportunity to discuss what was presented, and ask all the important questions which the Chairs forgot in setting up this symposium.
Stay tuned. We will be right back on-line in ten minutes, with Kerstin Matausch and further discussions on easy to read on the Web. Thank you.
>> SHADI ABOU-ZAHRA: Welcome back, everybody. This is Shadi Abou-Zahra. This is the second part of the easy to read on the Web on-line symposium. I'm making sure that we have Klaus back on and Kerstin and Andrea.
>> KLAUS MIESENBERGER: Yes, I'm back.
>> SHADI ABOU-ZAHRA: Okay. Please go ahead.
>> KLAUS MIESENBERGER: I would like to hand over, as far as we agreed that Kerstin will guide this discussion session. Right?
>> KERSTIN MATAUSCH: Yes, that is true.
>> KLAUS MIESENBERGER: Thank you, Kerstin.
>> KERSTIN MATAUSCH: You heard me, that is fine. Thank you, Klaus.
Yes, I just recognized, yes, welcome back to the last session today. Can you hear me better now? Okay.
During the last two hours, we heard a lot of interesting positions from the panelists. And parallel to this, we had rather intensive discussions on the IRC chat. And from this we collected a number of questions during this time.
And some of these questions could be answered, but some are still open.
Before coming to these questions, I would like to give you a short reminder on how to raise your hand. If you have a question, or if you have an answer to one of the questions, I will say, please remember you will get on the speaker list by raising your hand in the IRC chat.
You get on the speaker queue by using 41-pound key. You will get off the speaker queue by using 40, pound key. Is this right, Shadi?
>> SHADI ABOU-ZAHRA: Yes, exactly, on your phone keypad.
>> KERSTIN MATAUSCH: Okay, thank you.
So, we have questions regarding channel aspects, we have questions related to tools. For the beginning, I would introduce to go on with the question regarding channel aspects. So as far as I understood Clayton Lewis, he said easy to read is a more specialized service, it is more exotic.
So would it be possible or how could it be possible that we can modify individual opinions that easy to read, also on the Web, will no more be exotic, but will become a mainstream, part of mainstream information and part of a mainstream communication?
So, would any one of you give a note on this? Or would any one of you like to answer on this? Oh, I see Sean is on the speaker queue. Yes, Sean, please.
>> Actually, we have Mats Lundalv on the speaker list.
>> KERSTIN MATAUSCH: Thank you. Probably Mats is first.
>> MATS LUNDALV: Hello? Yes. One comment, we have had a discussion here on, I will not talk about the special issue of the text quality, but we have had a discussion about graphical symbol support.
And surely, we cannot get any common agreement about which graphical symbols to use, because there are very different traditions and opinions about that.
But what we can do is to unite around common infrastructures for providing lexical support in other modalities, including graphical symbol support.
That relates also to Clayton's point about providing lexical support with explanations of words and meanings.
And if we have a multi-modal lexical, lexicon support, multilingual and multi-modal lexical support, as an infrastructure for the Web and also for other network environments, that would be a great way to provide, to make this part of the mainstream, I think. Thanks.
>> KERSTIN MATAUSCH: Thank you, Mats, for this.
Sean, do you have also a note on this?
>> SHADI ABOU-ZAHRA: You may be looking at the wrong channel. Next we have (overlapping speakers) Clayton Lewis and Christian.
>> KERSTIN MATAUSCH: Now I see. Sorry. Clayton. Please.
>> CLAYTON LEWIS: On the question of how we can really get more people and more organizations to really involve themselves in this, I think it is interesting that there are cases of commercial organizations (Beep) doing a pretty good job on this, especially if we step away just from text and ask about who provides really simple interactions.
I think that the key there is the motivation. If you are trying to sell things, you are really paying attention to how successful you are in doing that. So this is a case of communication where people really are motivated to make the communication work well.
They are not just satisfied to sort of do something reasonable and hope for the best. They really want to follow up and make sure that people are able to get through whatever it is that they are doing.
In other situations (Beep) commercial motive isn't there to help us. This is why I said earlier, I think some change in how people think about things is needed. But perhaps a place to start is with agencies that have a specific need to interact with groups whose comprehension abilities are limited.
Of course, there are such agencies. We are trying such a, to get such a project going here, working with people who are responsible for programs that are (Beep) employments, and for people with disabilities. There are other examples like that. It could be that we do best with a relatively modest goal of trying to influence those organizations that have the strongest motivation to participate, and then hope that perhaps things spread out from there.
>> KERSTIN MATAUSCH: Okay, thank you, Clayton, for your answer. As far as I understood, it is about motivation and organizations to participate in the projects and to contribute this. Is that right?
>> CLAYTON LEWIS: It's observing that I think motivation really is fundamental. I'd say that is responsible for the lack of impact of the regulation that we have in the U.S., and it sounds like things are similar elsewhere.
If organizations aren't really motivated to communicate, what they are going to do is not going to make the investment to actually communicate clearly.
But where they are motivated, they will. Try to identify groups that have that motivation and working with them first is something that I think offers some return on invested effort.
>> KERSTIN MATAUSCH: Thanks for the motivating words. I see Christian on the speaker queue.
>> Hello. Could you hear me? Motivation is one of the key issues. I compare easy to read with what happened with blind people in the past.
On the Web, so what, and if you compare the two groups .... problems is much bigger than ones for blind people. If you talk about the benefits and why we do it, and we come on to who is benefiting from a clear and good communication, we might pay for the motivation we need.
Awareness raising towards design ideas to some extent, this goes with plain language, might be more than with easy ...
But anyhow, I think this is very important that we talk about ...
As has been said by Klaus, that it's important to provide .... in the processes of the Web design every day work. Otherwise ...
I stopped talking now.
>> I think we didn't understand everything you said. If I understood you right, you told us that the community, that the target group of people having a need for plain language (background noise) is much more larger than, for example, blind people, and that, yeah, could you repeat it and then clarify this, please?
>> Yes, that was just, I mean, to go back, in history, today nobody is, in accessibility it is commonplace that we provide alternatives to people who can't see and contrast people who see having the same problems and that was earlier not understood by many people saying blind people on the Web is not possible. Now we know it's possible.
Similarly today we talk about people with cognitive problems and many people say, cognitive people (distorted audio) on the Web, so what. I mean, we have to do a lot of awareness raising here, what the benefit is, how many people these are and benefit the general public. This comes, to the general public is more on the plain language level than the easy to read. So that was exactly what I was saying.
It is much more people in this case or this box than we have, for example, in blindness. From this kind of argumentation, I think we should pave the ground to make people aware that there is a lot of people out there who will benefit from this kind of services.
That was the first part of my comment.
>> KERSTIN MATAUSCH: Thank you very much for this, Christian.
We also received another question. It was regarding this: Couldn't one use the method of separating the tool and the content, like in CSS, also for easy to read on the Web? So would it make sense to separate tools, for example, checkers from resources being used, so for example, the grammar, grammar checkers.
And the question is could this be separated, since the tools can serve as an infrastructure? And could then be reused and transferred between languages?
Who would like to answer on this? Would this be possible? What do you think? I see Mats Lundalv in the speaker queue.
>> MATS LUNDALV: Can you hear me? Okay. I cannot speak for the grammar and syntax part of things.
But on the lexical level, we are actually working since several years on something we call the concept coding framework which is actually about separating the semantic definition from the actual representation, putting a layer in between.
So you define concepts in a concepts lexicon, and then link up different representations in terms of words in different languages, and symbol representations.
And that is exactly about this issue of finding a layer in between, so that you can manage alternative representation of meaning in a more efficient and cross language, and cross representation type in an efficient way.
In that respect, you definitely can do something there. And I think it's very worthwhile.
And to also link back to what Christian was talking about and what I was into before, if we work in that way, I think, and work towards this as a general infrastructure, we can have benefit for much larger groups than just the users with a need for easy to read content, because of cognitive difficulties. We have a large group of people struggling with a second language so there is a wider interest for this kind of language support.
I think we also can look at the countries where there is a large part of the population which is not literate, because there is a need for this kind of support and there is work going on in India and Africa, program, to provide services that are not as dependent on the language representation as we are used to here.
>> KERSTIN MATAUSCH: Thank you, Mats, for this. Would anyone like to contribute an additional note to, apart from Mats?
>> KERSTIN MATAUSCH: Yes, Clayton, please.
>> CLAYTON LEWIS: Yeah, a general comment on tools and the associated research agenda.
So as probably many on the call are aware, many of the tools we have for assessing readability are really correlation-based. That is, they don't actually measure comprehensibility. They measure things that are highly correlated with comprehensibility. And many people know Janice Reddish in the U.S. has written in particular about this important point.
And if you haven't heard it before, you might wonder what difference does it make? If I can measure something that is highly correlated with comprehensibility, isn't that perfectly good, and the answer is, unfortunately, it is not.
Those correlations are found in naturally occurring text, and they can be quite good there. But correlation can be very good for naturally occurring and misleading with respect to artificially created text.
In particular, if you create text with the intent of maximizing a comprehensibility score based on correlates, there are situations that show comprehensibility goes down rather than goes up.
Where I want to go with that is, our longer term agenda needs to invest more work in doing research on the actual processes of comprehension which is obviously a very difficult and long term kind of problem.
But if we look strategically at the issue, we are going to be in a lot of difficulty, and limited in what we can do, until such time as we have, for example, models that can use, that we can use to measure based on simulation, perhaps what the actual comprehension process is for something that we produce.
The tools we have now, as a rule, don't really give us that insight.
>> KERSTIN MATAUSCH: Thank you. I see Christian and Ginny. Christian, first.
>> I tried to be short. But I mean, I just read in the chat that [inaudible] mentioned again the U.N. convention and says that easy to read is requested in the U.N. convention.
This is a strategic and political statement now. I think, sorry, we have some problems with the audio. I think that it is important that the WCAG takes this into account. And I believe it is not sufficient to have only plain language issues on a level A in the WCAG, because that really fails then to support also a very big group of people with disabilities.
So maybe the work could lead to, let me say, a leveled approach which also takes action for people with IDD also on the first level, success level A. Thank you.
>> KERSTIN MATAUSCH: Thank you, Christian, for this clear position, which matches mine.
Yes, Ginny, would you please give your position to your answers, now?
>> Yes, can you hear me?
>> KERSTIN MATAUSCH: We can hear you quite well.
>> That is great. I wanted to support what Clayton had said about the difference between correlation and actual comprehension.
Most of the formulas and the tools have to make a balance between what is easy to count, and what really matters. Unfortunately, they often focus on what easy, what is easy to count. Plain language is much more than short sentences, and short words.
I think in the course of our discussion today, it's been mentioned several times how plain language matters depending on context, on audience. Plain language is also about the presentation of the information. I know easy to read cares a great deal about the space on the page, and the ability that people have to actually separate pieces of information.
The tools today don't take into account any of that. And the tools that were developed for print materials basically count very wrong on most Web pages, so that we really have to not rely on traditional readability formulas for understanding how to measure plain language.
We really have to include people not only in the development of tools, but in the evaluation of actual text.
And I really think that very important for us to remember. Thank you.
>> KERSTIN MATAUSCH: Thank you, too.
Next I see TS, on the queue.
>> There should be more than just a grammatical or sentence length analysis to really see what text, what makes a text comprehensible and readable.
I really would want to make a pledge for, a plea for doing empirical studies with real people to see what makes a particular text difficult to read and difficult to comprehend and difficult to apply, because most people are not reading for the sake of reading.
They are reading for the sake of wanting to do something with the text that they are reading. I completely agree with what Klaus Miesenberger earlier mentioned, that we might not think about a guideline that is about the text. The guideline may be which is more about the process, which could include user studies with a particular text.
That is what I wanted to contribute.
>> KERSTIN MATAUSCH: Thank you for this. So you think empirical studies are really necessary, they should include of course a target group?
>> Yes, sure, or maybe various target groups, because it could be that one, a text that is suitable for one group might alienate another group from a certain piece of information, for the same piece of information.
I think you should define in your guidelines what the test should, and what target groups needs to be included in the test.
>> KERSTIN MATAUSCH: Thank you for this also very clear position. Is there another one in the speaker queue? No, not so far.
Leealaura, are you still here, still participating? Probably you could contribute to this.
>> Yes. Can you hear me?
>> KERSTIN MATAUSCH: Yes, we can hear you.
>> LEEALAURA LESKELA: As I was just about to dial the number, what I would like to say is that I agree with what Clayton said and Ginny said, that about this perspective. I'm not sure whether this gives any new light to the discussion. But we have a service in the center here that is municipalities and other parties, organizations can send us texts and then we modify them to easy to read versions.
I understand that many of you do the same work in your organizations or where you work.
My practical, what I have the experience, is that when a text is logical, when it is thought through well, when the text level so to say is worked out fine, then it is just a mechanical question to transfer it to an easy to read version, to make it shorter sentences and change the vocabulary level to simple versions.
But in case when the writer of the ordinary text has not thought what he or she wants to express, what the main basic meaning of the text is, when the text level is poorly constructed, in that case, to transfer this ordinary text to an easy to read version is really a big challenge.
It is a difficult task. So I completely agree that we need to consider text as whole, not just sentences, not just words put together in a mechanical way. This is perhaps what I would like to contribute to this discussion. Thank you.
>> KERSTIN MATAUSCH: Thank you for this. Absolutely, I'm of the same opinion, the structure makes it more easier to transfer an information to easy to read, or even plain language.
Are there further notes on this?
If not, there is one more question, which refers to -- there is Patrick on the speaker queue. Yes, Patrick?
>> Yes, can you hear me? Hello?
>> KERSTIN MATAUSCH: Yes, we can hear you.
>> All right. Yeah. Well, as a content strategist and Web writer -- I'm not a standards person like a lot of people here. But I was reading the paper about the user groups regarding easy to read and plain language groups, and which to test.
I wonder how many groups to write for since language and communication is based on economy of means. I mean finding the perfect balance between the minimal amount of energy spent by both the producer and the recipient of any message in order for message to be understood.
There is a verbal cliff, if I may.
If I may use the term. We have to be aware of, any public organization, for example, in the U.S., the Internal Revenue Service, we need to use technical terms for advanced users like accountants, for example, in order to save, to be closer to the rules, the laws that are voted for.
And so it's not mistranslated into simplified versions for the general public. But it should also serve the general public with the simplified version, so in which I would recommend using the easy to read or plain language groups.
So my own numbers would be two different versions. I wonder do you have any take on this? What is your, do you see it the same way I do? For how many groups do you write for? Like with the script that you push certain version of the Website or another version of the Website, or -- so that is the question. It's a long one, but ...
>> KERSTIN MATAUSCH: Thank you for all this.
>> KERSTIN MATAUSCH: A lot of questions. I see -- one moment, Pia would like to answer to this and then Annika. Yes, Tia, please.
>> Same question came up at the Dutch, we will call tech demonstration or Internal Revenue Service, and they decided to create like two versions and present the information in more or less pyramid structure, where the text would open with a some size easy to read version of the regulation.
And for those who are more interested, they can drill deeper into the Website, and get the full regulatory text that for example accountants would need about tax issues.
That is kind of a combination. And speaking a part to two different user groups and write for those two user groups assuming those general information, general public who would be interested in the finer details would make the effort of understanding terminology that accountants also use.
>> KERSTIN MATAUSCH: Thank you, Tia, for this. Annika, you wanted to say something to this aspect?
>> ANNIKA NIETZIO: Yes. So I completely agree with what Tia just said, usually look at two different target groups for the information. One is the understandable version for the general public. And then for the most important information, there is a specific version in easy to read, which takes up only the main points and describes and explains them very clearly for the specific target group.
>> KERSTIN MATAUSCH: Thank you, Annika for this. I think there are two perspectives on this. One is to make summaries in easy to read for people who rely permanently on easy to read. And the other perspective is to provide full information.
So but I think there are two perspectives on this. I see Patrick on the speaker queue. Yes, Patrick.
>> PATRICK: Yes. Thank you, everybody, for the shorter answers than my question.
And yeah. Well, I want to serve both groups. But, well, is your take two different versions? Like one for general public, in which you use, you always use the easy to read version? And the other more technical one? That is the first part of my question.
And the second part is, yeah, how do you not duplicate the content so Google doesn't deprecate your content? Thank you.
>> KERSTIN MATAUSCH: Thank you, too, and also Annika. I see Ginny Reddish on the speaker queue.
>> Yes. Can you hear me now?
>> KERSTIN MATAUSCH: Yes, we can hear you.
>> I wanted to make the point that plain language works for everyone. And so it does not take multiple versions. Sometimes you want to do an easy to read, even simpler than plain language.
But there is no need to have a badly written document. What we really need to do is get everybody to write clearly. There is really good research that shows that when you follow plain language guidelines, everybody helps. It is not only illiteracy people that benefit from it, but high literacy people benefit from text that follows plain language guidelines.
So we are not really writing two versions when we write plain language. We really try to get everybody to do plain language.
That said, I know there is easy to read, and that is a further simplification. And also, there are times when you really want two different types of content on the same topic. So the U.S. national cancer institute, for example, has two versions of information on every type of cancer.
One is geared towards patients and their families. The other is geared towards medical professionals. Both versions are available to everybody. And they find that people on what we call a cancer journey start out in the one for patients and their families which uses less medical terminology, and is less in-depth.
And then as they become really really knowledgeable about their situation, they often go to the more technical version, and in the same way professionals go to the technical version for technical information, but read the patient's version to know how to talk to patients.
So that is another way of having two versions. But they wouldn't cause the Google problem that Patrick talked about.
>> KERSTIN MATAUSCH: Thank you very much, Ginny, for your comments, which are very very important, I think.
But we are running out of time. One more comment of Klaus, yes, Klaus, please.
>> KLAUS MIESENBERGER: Thank you very much. Can you hear me?
>> KERSTIN MATAUSCH: Yes, we can hear you.
>> KLAUS MIESENBERGER: Very good. I wanted to support what Ginny said before. I think we only need and we should make sure that there is one really well structured, well usable, well texted, well worded version, and getting rid of these very complex information which is on the Web.
Secondly, I think when we are in discussion of easy to read, and what I experienced with (Beep) a lot of talks unfortunately not in the study, what I experienced is easy to read very much goes into an individualized version or translation of information which is very hard to transfer from one group to the other.
And therefore, I might support an approach which says, as Ginny said, one version. Perhaps and also as Annika said, we have a summary in easy to read, that then way it goes down to very specific easy to read questions that we can define this as specialized services, be it on demand, that they are made available, but also in terms of feasibility, we should not and we must not expect that mainstream information providers might be capable of providing these very specialized easy to read versions.
>> KERSTIN MATAUSCH: Thank you, Klaus, for this.
In general, I would ask you to do the summary, because we are now starting to finish the symposium. Broadly, you could make some (typing) Words for the conclusions.
>> KLAUS MIESENBERGER: Sorry, was this for me?
>> KERSTIN MATAUSCH: Yes, Klaus.
>> KLAUS MIESENBERGER: Yes, now I got it.
>> KERSTIN MATAUSCH: Conclusions now.
>> KLAUS MIESENBERGER: First of all, I would like to say thanks a lot to the Chairs. Thanks a lot to the panelists for all the input of the papers we got.
I think we have a very well elaborated and a good bunch of work to come up with a report of the symposium which I think would really help and impact on the work and the research and development group.
Thanks also to the colleagues from WAI, Shadi in the background, giving a lot of support during the symposium.
It is hard out of the (Beep) lot of information to make a conclusion. But I think as said before, plain language oriented towards better general use ability and readability of information is a key issue, which is not only for our target groups, which is beneficial for the owners of the Web which is beneficial for the authors of Web content, and of course which is beneficial for all the users of the Web.
And, also, there is still a strong need for tools helping to support plain language, helping to support easy to read, and of course, we need very well elaborated services, well trained professionals who could then provide the level and quality of information which specific user groups need.
Once more, let me say, thank you, and by this close our discussion today. I'm looking very much forward to get feedback from you, and in the future, perhaps to see progress in easy to read plain language on the Web. Thank you very much.
>> SHADI ABOU-ZAHRA: This is Shadi Abou-Zahra. I'd like to also thank Klaus, Kerstin and Andrea for putting together such a great session.
We have had really good contributions and good discussion. I do encourage people to, if you have follow-up comments, input, thoughts, please send us, there is an E-mail address posted on the Web Page of easy to read symposium.
There is a list pointing to the publicly archived meeting lists, so we can reference your comments and your input. This will all be considered in this symposium report, the research report will be the outcome of this symposium, which we will announce but also send to you when the draft is available for you to comment on.
I'd like to mention that today is also the International Day of Persons with Disabilities. And so it's I think a particular good day to look at people with cognitive disabilities who, where more research continues to be needed, research and development.
And this is the focus of the research and development working group of the Web initiative that looks at research in the field to help inform standard validation and development.
If you are interested in learning more about this group, please feel free to contact me or Simon Harper, who is the co-chair of the working group. We would be happy to provide you with more information about how this group works, and how you can get involved in accessibility research.