25/10/2008 by nicola.

I was about to spend part of my weekend trying to figure out some of this, when I saw Dave Pollard’s post about connectivism & the week’s subject being about complexity, connection and learning. What ! So just to be clear, there are around 2000 people learning about this, and I am trying to figure out how the hell to visualise a wiki structure by myself. I mean 2000 people, there must be people with amazing technical and design abilities studying / talking about all this - and what, imagine if you have even 4 people, one with Java, one with design, one with wiki and one with - well whatever -it would take - 1 hour, 2 hours top to figure it out ?

I’m going to go and rediscover my weekend ! There must be amazing conversations and ideas flowing out from their experiences right now, I hope that if anyone out there knows of anything they could take maybe 5 seconds out of their day to send me a link or something and if not, so be it, but I don’t see it as a productive use of my time to spend any longer blogging about wiki stuff, happy to leave it to greater minds

Posted in wiki, visualisation | Print | No Comments »

20/10/2008 by nicola.

Got distracted by still trying to embed dynamic ‘things’ like a Wordtree visualisation on wiki pages - think I’ve found a way to do which haven’t done yet -on Google Sites wikis. Just to make it confusing for anyone unfamiliar with Java (programming) and JavaScript (browser scripting) they are not the same, but the Java applet visualisation in Many Eyes can be embedded using JavaScript. You can use with Google Gadgets API as per this explanation on Google groups by embedding in an iframe which lives like a floating box on top of your regular web page. iframes are not great from an accessibility point of view, but you can use them on various wikis. Don’t have time to look at an API right now but will return to…

Whilst I’m going completely off topic, if you ever want to know more about JavaScript, there’s a nice friendly tutorial that explains it using wild bears.

Profiles
If you have an enterprise wiki solution, there is a possibility that your wiki will be talking to other applications you use, maybe all linked so the chances of having more information other than your name is more likely. However for public free hosted wikis, it is more likely that the profile information is not likely to be as detailed because it partly depends on the person who has filled it in - with the multitude of things we sign up for, filling in profile information each time into different boxes in forms is less and less appealing. Or in some cases it is little more than a contact details form e.g.

With some wiki software, they give you the option to add tags to your profile - this could be about areas that you might be interested in. So if someone is using a wiki for the first time and would like to find other people as well as content with similar interests, the tags can be really helpful. You could then create visualisations of tags, in the same way as many of the delicious add-ons that have been created e.g. in Malinka Ivanova’s great post earlier this year.

However, this is starting to stray into social graph territory which is way beyond simple visualisations of who is doing what on a wiki and the additional visualisation of other areas in which a person might be social, could be distracting rather than helpful - will it actually make a wiki more social / does a wiki need to be more social?

Bearing that in mind, will attempt to do experiment 3 - soon. This series of posts is as usual, my attempt to throw some ideas out there - I know a whole bunch of stuff about web pages and wikis, I’ve been playing around with charts, formulas in spreadsheets (can’t really do anything in egovernment without being buried occasionally spreadsheets !) and some visualisation stuff for several years but do not consider myself to have expertise in either of these when it comes to extraction of data of any kind. I’m kind of hoping / dreaming that anyone who comes across these posts with more experience in a particular area can point me in a useful direction…

Likewise for experiment 5 when I try and put some of this into a 3D environment, I have until now purposely avoided things like Lindenscript (and no room to think with all the other programming languages I am attempting to do stuff with) and I have not built anything other than a media and web page ’screen’ in Second Life so far. I’m really hoping someone else has already found a way, there is a wikiHUD for Second life wiki which may be a starting point.

Posted in wiki, visualisation | Print | No Comments »

19/10/2008 by nicola.

Based on my experience with wikis, pages can be created in two ways - using something like an Add Page or Create page button, or by creating a hyperlink - such as tikiwiki and from what I remember, tiddlywiki (also open source) uses the same kind of idea.

But first a note about web page design. If you are doing web design and you meet a client for the first time, one of the most useful things you can do is sketch out (digitally or on paper) like a flow chart - the navigation - as you get ideas - because once you get beyond your first six ideas it can quickly become complicated trying to work out what needs to link back to where and how. Lots of web editing and web publishing tools have this built in, so you can see your web site structure growing as you add possible pages.

With wikis, the game completely changes because most wikis don’t have built in options for visualising possible structures and with lots of people as editors, they just naturally create pages as they get ideas or attend events or whatever and suddenly you end up with lots of content which may well have internal hyperlinks within the page content itself but other than a list of “All pages” or something like that, it is difficult to understand what is going on inside the wiki as a whole. Also people want to find information quickly and if they can’t visualise where it might be, they will need to search (another web page opens on their screen) or ask and wait for answer - and it would be useful to reduce the need to do this. Users / learners want multiple paths through content and providing ‘maps’ is one way for anyone that would like to plan a route

Once wikis start growing, it might be helpful if you could visualise how it has grown and the patterns of links between pages, it might provide some insight into the ways people like to organise their thinking. In a recent Google Texttalk, a visual wiki was demonstrated by a university in Auckland:

(recommend high quality view, approx 33 mins)

You can see in action E.g. Thinkbase (Thinkmap+freebase), Thinkmap+Confluence, Thinkmap+Mediawiki shows visualisation of content in a map on LH side based on content of words in RH side - not an overall visualisation of wiki as in pages etc It doesn’t include mapping of comments and uses Java.

Most visualisation tools that I have played with seem to display with Java or Flash. Confluence which is one of the best wiki software I have used (Atlasssian, Stewart Mader, Wikipatterns etc) is not open source but has an open source version available if you can demonstrate that your organisation is already using open-source tools etc Media wiki which is open source has a lot of options for integrating with other tools, there are lots of developers making that happen. Don’t know about Freebase.

The completely awesome Chris Harrison has done some brilliant visualisations from Wikipedia too, this is the kind of thing I would like to have achieved (but couldn’t - just don’t have necessary understanding to do yet). Just like with a tool like Visuwords with each wiki page title appearing and related pages or something like that. There are several other wikipedia visualisation examples on 95 and 96 of the meryl.net list but I haven’t had a chance to look at those yet.

So onto the experiment. I used this Wikiversity page and filtered it to 500 results. Then cut and pasted the page titles as text only, into notepad then put in open office calc sheet, manually pasted user data into second column and redirect pages into third column then saved as CSV. Deliberately not trying anything using formulas yet btw, trying to imagine you are a user with no experience of spreadsheets (!)

Tried uploading into Swivel, but it just displayed as a table (I might have done something wrong) so no good. Had a look through various visualisation tool lists but couldn’t see anything quite appropriate so returned to Many Eyes. Tried various visualisation options such as Treemap, but couldn’t get the data to display. So was reduced to going back and using WordTree again, which is useful but not as far as hierarchical structure of pages.

Tried using chart tool in open office calc sheet but no different - 500 rows makes it unreadable, which I think is why it didn’t work in Many Eyes. So without pulling all the data in a more meaningful way than just cut and pasting text (or even if I had selected the text as hyperlinks, same result for now), so if you are a user with no experience of using formulas or manipulating data in spreadsheets, it is not going to be that easy - I have yet to find a way to get it display in a map format of some kind. You could reduce the list from Wikiversity to 20 or 50 and try again but I have had enough for one day

Experiment 3 will look at possibilities of finding and visualising relationship data, by looking at profiles and whether its possible to pull relevant data from a ‘web form’ then display it. I’m not confident what can actually be achieved with this, started having conversations 1, and somewhere in the middle of 2 on InternetTime Ning community last year. Experiment 4 will go back to the content again, I’ve had a very quick look at some of the features in open office spreadsheets etc so will be closer to the kinds of things Tony Hirst mentioned in his blog. Experiment 5 , if possible will be to try and work with some of this in 3D.

Posted in wiki, visualisation | Print | No Comments »

18/10/2008 by nicola.

Tools used: Pbwiki, mindmaps - various, Tagcrowd, Wordle, Many Eyes - all free tools (to some extent, see mindmaps below for further info)

Wiki have been playing around with today is pbwiki - due to having text and some tags that I could use. I looked at 4 mapping tools today - mindomo, mindmeister, mindmymap and cmaps - very briefly and want to state that there may be ways of doing these but I couldn’t find them personally today.

I wanted to know if it was possible to just cut and paste text from a webpage - i.e just lifting it straight from the words you see on the wiki page i.e. put it into Notepad and remove any formatting weirdness then paste into one of these tools. Or the wikipage from its markup, or go further and select the source code (html code that the web page is written in)
e.g. IE 6 & 7 select Page and View Source,
Firefox= View then Page Source,
Chrome= Page icon, then Developer then View Source,
Safari= View then View Source
Opera= View then Source or Ctrl+U
Then as with just copying text direct from the web page, stick into Notepad then paste into a map using one one of the tools.

This is a backwards way of using a mind or concept mapping tool because the text had been already created on the wiki, but I was curious as to whether you could find key concepts, common themes. There was no way of importing or pasting - in some of the tools you can attach files (but not in free versions as far as I could see) but I don’t know if these just live as an attachment ‘icon’ on the map and therefore not searchable.

So then thought maybe tags - find the tags from the wiki and put these into a concept/mind map and play around with the features to see what kind of visualisations could create - but with this particular wiki there were only 12 tags created so far that were related to the wiki content, so didn’t.

I have used tagcrowd in the past - I know I can embed it into wikis because I’ve done this when I was using a wiki which didn’t have a tagcloud feature, depending on the wiki - unfortunately most wikis currently have different ways of coding things when you put in html code (like the kind of code you grab for widgets, gadgets, badges etc and paste into your blog or site). Wiki-Creole is a project attempting to try and create common wiki coding for all.

Anyway, with tagcrowd (its like Wordle but less display options, but you can show the frequency of a word which is useful). I just copied and pasted directly from the web page (without going to the source code) into the paste box in tagcrowd and selected the following(please click on images to enlarge):

This was the result:

I then tried inserting it, using the html code provided, back into a pbwiki page. It didn’t work - well it wouldn’t add it as html (pbwiki magic option) separately so I tried sticking it in by clicking source and got this:

To cut a long story short, you would need to style it yourself in order to make it look as interesting as a regular tag cloud, ie the more popular words bigger font etc On a different wiki in the past I have not needed to do this, so it may be possible with other wikis, I don’t know yet.

So, went to Many Eyes next. Many Eyes is amazing software where you can upload all kinds of data including text and create visualisations, any time you use it, it will be made public. I pasted in the same text that I had used for tagcrowd and found some interesting visualisation features.You just click visualize and scroll down through all the options.
Tagcloud - it didn’t show the 500+ but you can search for tags.

I found the Word Tree useful in terms of visualising the different structures - it gives a much better idea of how people might have attempted to organise their thinking. This wiki is being developed by politics students and is about the European Union, so I just put in the word European. Then clicked Publish to publish it - if you look under the visualisation and choose share this, it gives options to share as a static image or as a live visualisation.

There are so many other amazing features with Many Eyes, including commenting on and watching visualisations so they really are such exciting, living documents. I can’t embed this back on pbwiki currently because their 2.0 version does not support embedding scripting such as JavaScript, but I think the Word Tree definitely has potential to pick up ideas about the content of that wiki page.

UPDATE: This is the same data as a Wordle too. I have tried entering the html code to display it as a plugin (pbwiki magic), into the source code but it doesn’t show up. So I took the image URL from wordle.net and tried inserting as an ‘image from URL’ and that didn’t work either.

Experiment No2 will be looking more at the overall visualisation of a wiki structure rather than on the content, i.e. once you get beyond 10 pages of a wiki, how can you start to visualise how it has grown. Am not sure at this point if can do with all open source tools -a couple of ideas but not sure yet.

Posted in wiki, visualisation | Print | No Comments »

18/10/2008 by nicola.

Yet again Tony Hirst has done some magic on OUseful with data, Yahoo Pipes (I created my first pipe after reading through Tony’s example whenever it was - a year, 2 years ago?) A thousand thanks again to Tony for explaining everything so clearly as ever with achievable steps that I can follow. I could sit there for years and it would probably never occur to me that I could find an import html function in a spreadsheet.

Brilliant - so much to go and play with now - will put mobile stuff on hold for a bit. If you want to look at mobile learning stuff, check Ignatia webs, she has already started to do some great thought-provoking write-ups from Mlearn. Indeed if you ever want to look at mobile learning stuff, Ignatia webs is one of the wonderful places to start, in fact 8 out of 10 mobiles said they like what they see there.

Posted in wiki, visualisation | Print | No Comments »

13/10/2008 by nicola.

Have removed personal info and checked re ’sensitivity’ so please find attached: Wikis and Cognitive Processes. There are not any plans to release this formally, this was designed as an informal paper to continue discussions around some of the wikis at Surrey, one in particular sparked this and they wanted within a week to start making decisions. Due to semester start, workload etc, I only went to sources where I could find info really quickly so there are not a vast amount of references/sources. It is not easy to find info around these issues, specifically related to wikis. Its not easy to think your way around these issues. Also, in my case its not easy to attempt to pull together thoughts in some kind of coherent way about these issues either

Any comments, thoughts would be most welcome !

Posted in wiki | Print | No Comments »

02/10/2008 by nicola.

Have just completed a first draft of a paper designed initially for internal use at Surrey - the scenario is that some students who are using one of the wikis at Surrey may be finding it challenging to access, view and process information from a wiki and represent their own ideas and views as part of a group contribution. The specific nature of these issues is not and may never be known unless the students wish and are able to disclose the information. In the wider group there are also others who have previous experience with wikis and one had edited articles in Wikipedia.

The paper aims to look briefly at how cognitive processing may influence and be influenced by collaborative editing on a wiki, brief discussion of whether a universal design approach would be helpful due to the possible diversity of cognitive processes and challenges , i.e. is there any such thing as a normal student?

Additionally some suggestions about recommended actions to help students generate knowledge on the wiki, expanding, amending their own ideas, improving their editing / writing and negotiating their way through complex group conversations.

If anyone has a specific interest in this area and would like to take a look, would really welcome any comments, please could you supply some form of contact details, thanks !

All references for this paper are on my delicious links via wiki-usability

Posted in wiki | Print | No Comments »

18/09/2008 by nicola.

A million wikipedias

Would like to thank the following p e o p l e and many other very wise people whose thoughts on wikis and collaboration - have helped shape my own thinking as I have been talking about wikis with people for a few years - and in the last 24 hours too as am preparing for some wiki training which am about to deliver. Have never done wiki training before - very exciting - have produced guides and lots of chats with individuals and small groups, (kind of took up residence in wikis in 2006) but not often you get the chance to go and discuss wikis with 80 students.

So what are wikis anyway - are we in reality with all the different wikis out there, creating millions of online sets of wikipedias ? Does that matter ? What do we use wikis for?

No one could create a wikipedia - that would be like someone writing the whole web about everything and everyone in existence and has even existed. Having said that, I am currently using one wiki in a most unwikilike way - as my online notebook because I find it much easier for linking and putting together other info quickly, that I can return to whenever - and figure that whilst on a wiki, don’t have to bother with sharing certain bits and/or permission levels - just make it public - whoever stumbles across it - they may or may not find anything of interest and can contribute / edit or whatever (I do backup occasionally in case someone deletes the lot, but may not even bother with that in the future, its its meant to be…)
I’ve come across a few people who are also using wikis for this purpose too. In my case its just about a wish to make information and ideas public and my personal content creation preferences.

What about command and control? I joined citizendium but mostly due to lack of time initially, I didn’t contribute anything. Then somewhere along the way I decided that citizendium would be better as a widget on wikipedia - rather than a rewrite of everything already on there. If there is information where people feel sufficiently moved by its contents in terms of whether they think it is appropriate / relevant / accurate and/or is about / relates to them in a personal way - but don’t wish to either edit or discredit the information themselves - could use something like citizendium within a widget on that page.

If it was a widget, it could have links to other sources of information that provide alternative viewpoints to the one on wikipedia. The widget could also highlight that there is not only one viewpoint that is accepted as reliable or ‘truth’

As in the great words of Kahlil Gibran (written 1923)

Say not, “I have found the truth,” but rather, “I have found a truth.” Say not, “I have found the path of the soul.” Say rather, “I have met the soul walking upon my path.”

I can’t wait to hear what the students think of wikis and see how they are going to make use of one, how useful they will find it - can they discuss, debate, will they be adopting different wiki patterns and antipatterns or wikigardeners, how they will feel about content and ownership etc

Posted in wiki, web design | Print | No Comments »

16/09/2008 by nicola.

I’ve got a todo for Surrey which not quite getting round to due to semester start, it is looking at ways of visualising research, hopefully in the near-ish future from wikis but initially will be from other databases / web pages / any other sources (could also be spreadsheets and documents too). Steps may be:

1a) Depending on how complex it is to pull the data from these sources, would like to create visualisations - Many Eyes seems like a good starting point, due to ease of text upload and variety of visualisation options available.
b) would also like to pull relationship data - not just project info but profile data too - SweetTools has a huge list of possible ways of pulling data (most of which look technically beyond me personally)
2. Display this visualisation in a 3D environment and also on a wiki. I have started to mess around trying to insert mindmaps (mindomo, mindmeister, ekpenso so far) as collaboratively dynamic editable objects onto wikipages without much success so far (going to try Google Sites next, this may work in which case Google is entitled to take more of my soul). I haven’t tried any graphs, charts - maps you can insert but as per above I have not found ways to make these very collaboratively editable as objects within themselves.
3. Have a wiki in 2D and a linked visualisation and/or wiki in 3D environment which can be edited / updated either via 2D or 3D, depending which dimension you or your avatar find yourselves in.

I’ve been looking at the Studio Wikitecture project - I remember seeing something about this ages ago but now in more detail, the video is amazing. You can also visit the completed project in-world in Second Life. I know there are possibilities using search engines in 3D / visualisation as a search engines, but this is a great concept too. If anyone interested in being involved too, please do let me know…

Posted in wiki, visualisation, virtual world, Technology | Print | No Comments »