Sasan Salari's blog

Updated WebCT Authentication plugin for Elgg 0.8

Wed, 22 Aug 2007 10:07:30 -0700

Elgg 0.8 introduced some small changes which required modification of the WebCT-authentication plugin.

We have just released version 1.0.2 of the plug-in, which is available from the download site.

The plug-in allows Elgg to authenticate users against the WebCT Vista (or CE6) database, thereby eliminating the need for administrators to set up duplicate accounts.

More information can be found on this page.

Update to Elgg-0.6-MediaWiki package

Tue, 23 Jan 2007 04:59:56 -0800

It was pointed out to me by a user of the Elgg 0.6-MediaWiki integration that a file was missing from the distribution zip. The file (error.php) was part of the Elgg 0.4 - MediaWiki integration package, so if you upgraded in place, you would not have seen the problem.

The omission has now been fixed, so if you experienced any problems with the integration, please download the package again and update your installation with the additional error.php file.

Apologies for any inconvenience and puzzlement this may have caused.

Elgg 0.6 Integration with WebCT

Thu, 09 Nov 2006 19:48:31 -0800

The integration between Elgg 0.6 and WebCT is now available on our download site. The WebCT Powerlink is identical, so anyone migrating from Elgg 0.4 does not need to update anything on the WebCT side.

The integration itself remains the same functionally - the Elgg-portion of the code has just been updated to take advantage of the database abstraction layer that was added for Elgg 0.6.

Authenticating Elgg users against WebCT Database

Wed, 08 Nov 2006 22:22:36 -0800

A few users of the Elgg PowerLink have suggested that it would be nice for their students to be able to log into Elgg directly, without first having to pass through WebCT.

We have just completed an Elgg authentication plug-in which addresses this request. When users access Elgg's login page and enter their username and password, Elgg's internal authentication is bypassed, and a web-services connection is made to the WebCT server instead.

If the username and password match those stored in the WebCT database, the user is logged into Elgg and is presented with their Elgg homepage.

For more information and download instructions, please see this page.

As with the Elgg PowerLink, this authentication plug-in is provided free of charge. If you have suggestions on how it could be improved, please let us know.

ElggSpaces Hosted Solution now available

Mon, 06 Nov 2006 10:58:21 -0800

Curverider, the company founded by the creators of Elgg, have just launched a turn-key hosted solution for Elgg at http://elggspaces.com.

This is a much-needed option for those institutions who don't have the technical expertise or the staffing resources to host Elgg for themselves.

Since the service is being run by Elgg's developers, users can expect all of the latest bug-fixes and features to be available as they become available.

MediaWiki-Elgg Integration updated for Elgg 0.6

Tue, 15 Aug 2006 09:44:01 -0700

We made some small changes for the MediaWiki-Elgg integration to support Elgg version 0.6. The original version was coded against Elgg 0.4.

Specifically, Elgg 0.6 supports a database table prefix for the Elgg tables, which is now a configuration option for the integration.

MediaWiki-Elgg Integration Posted

Fri, 28 Jul 2006 12:13:32 -0700

I have just posted the MediaWiki-Elgg integration for download off the Aperto download site. More information about the integration can be found here.

Comments and feedback welcome.

WebCT Impact 2006 Wrap-up

Wed, 19 Jul 2006 12:25:41 -0700

I am back in Vancouver after an eventful week at the Annual WebCT Users Conference, held in Chicago this year. It was a bittersweet event for me, since it was the last official WebCT-only conference -- all future ones will be joint Blackboard/WebCT events. I was able to reminisce with a few old-timers about the past 8 annual conferences, and the great energy and enthusiasm the attendees and the staff have brought to these events.

Ok - enough of that. Let's get to some the major take-aways for me:

Blackboard's Beyond Initiative

This initiative was announced in March at BbWorld, and originally focused on 4 areas: Global Learning Objects Respository, Social Networking, ePortfolios for Life, and Outcomes Management.

In the opening keynote by Matt Pittinsky and Michael Chasen, a fifth element was added, which was Student Centered Learning. This was of particular interest to me, since it happens to be the exact title of the handout that I created for the WebCT-Elgg integration, and which was an insert for the conference bags. It is good to see that everyone is thinking along the same lines.

Blackboard is just starting on these initiatives, and clearly stated that they would be multi-year projects. While real details were not yet available, I hope that these central sites will have a trickle-down effect to the product lines, so that web 2.0-type functionality can be taken advantage of directly within the course environment right across the institution.

Blogs, Wikis, RSS and Podcasts are everywhere

There was a huge buzz on these topics at the conference, with many sessions and pre-conference workshops touching on one or more of these. Most of the users take advantage of centrally-hosted commercial sites, such as blogger, blogspot, del.icio.us, so I took the opportunity to mention Elgg as a locally-hosted option in many of my conversations.

Dr. Helen Barrett's Closing Keynote on ePortfolios

Very good presentation touching on a number of key points (Download slides here).

The take-aways for me were:

Concept of Working Portfolios vs. Presentation Portfolios
Is Portfolio to be used as an Assessment of learning (summative), or as an Assessment for learning?

The Closing Video

It would not have been a WebCT conference without some comic relief from Utah State's Kevin Reeve and Marc Hugentobler. This time, they wandered around the conference with a camera crew and captured some great moments. The sailor suits for the conference party at the Navy Pier were a nice touch.

At the WebCT User Conference in Chicago

Mon, 10 Jul 2006 13:13:48 -0700

Hello WebCT-clients - I am currently attending the WebCT User Conference in Chicago. If you are interested in having a chat about Elgg and the WebCT PowerLink integration in person, please feel free to send an email, with potential times that work for you.

I will also be demonstrating the integration as part of the following conference activities:

my talk on Thursday morning, July 13th, at 9:00 am in Room Michigan A
a demonstration in the Vista Developers Network booth in the exhibit hall on Thursday July 13th, from 10:45 to 11:15 am.

I am looking forward to catching up with you all in Chicago.

Sasan

Elgg Outside of Higher Education

Tue, 20 Jun 2006 14:23:32 -0700

Over the past few weeks, I have been involved in a project using Elgg in a non-academic environment.

Vancouver is currently hosting the World Urban Forum, an international event on global urban sustainability. Working with teams from UBC and Opn Design, Aperto helped set up two sites to support activities related to the event:

Earthblog.ca is a site to promote dialogue around social, environmental, cultural and economic issues facing the Greater Vancouver region.

Elgg's main homepage was branded with logos, and the entry page of Elgg pulls in the most recent postings of four key bloggers, as well as a 'hot comment' (as designated by one of the site admins periodically), and the most recent comment to the site.

Most of the Elgg functionality was disabled for visitors, except for the profile tool. The official bloggers have access to all of the Elgg functionality.

Another key part to this site was the single-sign-on integration between Elgg and MediaWiki. All login is handled by Elgg, and the session is transferred to MediaWiki if a user wants to edit a particular Wiki page.

We will be releasing the Elgg-MediaWiki integration to the community shortly, as a number of other Elgg users are looking for this functionality. More on that in a future post.

The second site Aperto worked on is GUSSE, a social site to collectively discuss, review and apply the best ideas for sustainable cities.

The integration here involved a customization to the Elgg entry page again. Registration and log in is being handled by the social tagging system opntag. Elgg reads the opntag session information and creates users in the Elgg system. GUSSE maintains all Elgg core functionality.

It has been interesting applying Elgg to a non-academic environment, and I hope that these examples help showcase the flexibiliity of Elgg, and the ability to customize the look and feel of the application.

WebCT-Elgg Integration Flash Demo

Wed, 19 Apr 2006 12:04:35 -0700

After getting a number of follow-up requests from the WebCT European Conference in Edinburgh, I decided to create a short Flash-movie which demonstrates the WebCT-Elgg integration.

The file can be found at:

http://www.aperto-elearning.com/files/elgg_demo.htm

The demo is best viewed on a 1024x768 or larger display, to minimize any scrolling.

For those interested, a PDF version of my presentation slides are also available for download from the conference site at:

http://gromit.webct.com/webct_europe_2006/Tu/Salari.pdf

Del.icio.us Tagging

Wed, 22 Mar 2006 20:06:08 -0800

Thanks to a note on Brian's blog, I attended a session by HP Researcher Scott Golder on his study of tags and social bookmarking on del.icio.us.

The talk was based on a paper which is due to appear in the Journal of Information Science.

Scott first gave background information on issues encountered when tags are defined by a community (ie. folksonomy vs. taxonomy):

synonymy: using different tags for the same thing (tv vs. television)
polysemy: a tag may have related senses (he gave the example of a window - either the hole in the wall, or the pane of glass)
homonymy: a particular tag can mean different things, depending on context (eg. speaker, chair, apple)
the 'basic level' problem: describing 'something' is related to the person's level of experience with that particular element. For example, a picture of a dog may be 'dog' to most people, but 'beagle' to a veterinarian.

In the analysis of the tags used to describe the bookmarks on del.icio.us, Scott found that they fell into 7 categories:

Identifying what or who it is about: topics of the bookmarked items. Tags consisted of common or proper nouns.
Identifying what it is: the kind of thing being bookmarked (eg. article, blog, book)
Identifying who owns it
Refining categories: these refine or qualify existing categories
Identifying qualities or characteristics: adjectives used according to the opinion of the bookmarker (eg. funny, scary)
Self Reference: identifying content in relation to the tagger (eg. mystuff, mycomments)
Task Organizing: information related to a task, used to group related information together (eg. jobsearch)

From this list, it can be seen that some tags are general, whereas others are very personal to the tagger and their experience. From an analysis of tag order, it appears to people use generally meaningful tags first (and these have the highest freqency across bookmarks), and more personal tags later. Therefore, analysis of the first few tags of a bookmark should give a good idea of the general topic of that link.

Scott also found that users' tag-lists varied greatly, and were not related to the number of bookmarks a particular user had. In one case, a user's number of tags increased steadily as their number of bookmarks increased, while another user's number of tags stayed level. This points to definite personal styles for tagging.

Another of his points related to usage of new tags. For example, I may have been doing some research on Africa, so I have tagged all of my bookmarks with that tag. However, a few months later, I decide that I need to be more specific, and add a country tag as well. This additional level of classification poses the following problems:

I cannot search my old bookmarks for instances of the country tag
Adding a country tag to my existing bookmarks (which could be in the hundreds) could be extremely onerous.

Finally, he found that tag proportions over time tend to be stable for a particular bookmark. This means that, after approx. 100 bookmarks for a particular URL, each tag's frequency is nearly a fixed proportion of the total frequency of all tags used. So for example, a particular URL may have the tag 'dog', which will have a frequency of 30% of all tags used for that bookmark, regardless of how many bookmarks are added and tagged for the URL.

The implication from this result is that "after a relatively small number of bookmarks, a nascent consensus seems to form, one that is not affected by the addition of further tags."

Some of the questions I am asking myself after the session are:

Do I tag for myself or for others? In the case of bookmarking, it is mostly for myself, but what about tagging my blog entries - is it so I can find them more easily later and selectively filter, or do I tag to make it easier for others to find my information? And if I do tag for others, do I tag with a specific audience in mind, and does that intended audience then define to some degree the tags that I use?
My own set of tags is evolving over time, based on my experiences. What is my responsibility for updating the tags of any previous content that I have tagged, should I decide to split a particular tag up into 2 or 3 different tags? On the one hand, the thought of inconsistency makes me uneasy. On the other hand, maybe I just have to accept the evolution of one's tag set, and consider the older tags as a snapshot of my state at that time.

How much structure for social software?

Mon, 20 Mar 2006 23:30:42 -0800

How does an institution successfully increase and scale usage of social software across campus?

Any institution that has deployed a VLE for a number of years likely has a significant amount of usage by faculty and students (I can't recall the exact Gartner figures, but for North America, I believe around 50-60% of courses have some type of online presence.)

The good news about that number is that online learning is now mainstream. The bad news is that much of the online presence is very limited - maybe a syllabus, a couple of web pages, possibly some e-mail support. In other words, not very engaging.

Having all of these courses online now provides an opportunity to build on that user base, and to introduce them to new technologies, such as social and student-centered tools. As I wrote in a previous post, the first step is to make it as easy as possible to get started. If users are already in the VLE environment, that means automated sign-on and navigation between systems.

Once users are in the new system, what are they to do next?

Most social software is very flexible and unstructured, exactly the characteristics that make it so attractive in the first place. What then is the best way to expose students and faculty to these services?

I am reminded of two bits of information here - one from a few weeks ago, the other a few years old:

In Edinburgh, at the ePortfolio roundtable, one attendee described their University's experience with two systems. The first system was an 'empty shell model' which allowed students much flexibility in how they decided to use the features. The second was less flexible and provided more structure through the use of web forms. Students preferred the second system, which received significantly higher usage, because they simply did not know how to use the more flexible toolset.

My second example is from the early days of WebCT. Back then, when a new course was set up, the instructor simply saw a blank page when they logged on, along with some action buttons at the bottom of the screen. Again - the instructors had a lot of flexibility on how they decided to structure their courses, but most where overwhelmed with the choices and had no idea on how to begin.

As these tools are rolled out across a campus, some initial structure is inevitable, and there are a number of possibilities:

'soft structure', in the form of documentation, training and best practices. The issue here will be the ability to reach a significant number of users with existing resources.
simplification or elimination of features - it may be possible to hide or disable some of the software features, to make users comfortable with a subset of functionality, and gradually re-introduce functionality in stages.
templates and context - providing users with pre-populated content based on existing data, auto-populating communities that mirror academic structures, guiding users through the initial set-up steps online.

In all cases, the institution has a role in defining how these tools are to be used, and has the choice of suggesting vs enforcing. However, care must be taken that any enforcement of structure does not diminish the utility of the tools over current technologies.

Probably the most important point: users must be able to use these tools successfully in their initial attempts. It is much easier to introduce more complexity over time than to reverse a negative first impression created by an overly difficult system.

Edinburgh Conference Experience

Wed, 08 Mar 2006 22:33:36 -0800

I am finally getting around to posting a summary of my experience at the WebCT European User Conference in Edinburgh last week. Blogging is obviously not yet second-nature to me - according to some feedback at the Northern Voice conference in February, it seems to take new bloggers a few months to get into a rhythm, so there is still hope for me.

Overall, the conference was a great experience (as all WebCT user events tend to be). The program is very user-driven, with many great presentations and insights from the participants. I particularly like the European event every year, as it brings together so many different cultures and voices, while still maintaining a cosy and manageable size (roughly 350 participants). The scheduling was also good, with decent breaks between sessions and at lunch to allow for unrushed interaction and discussion.

This was also my first event as a non-WebCT employee, so I spent a fair bit of time explaining what I was up to. Before the conference, I was actually not sure how my work would be received. I am certainly excited to be working on the Elgg integration, but would those in the WebCT world share that enthusiasm? Are people thinking about incorporating student-centered learning into their teaching practices? Is there an interest in using blogs, social linking, podcasting, or are these just buzzwords being bandied about by the early adopters?

The answer I received was a resounding 'yes'. I had a great turnout to my presentation, and approximately one third of the audience had experimented with blogging, podcasting or wikis in their teaching, which was very encouraging. I was also happy to see a diverse set of European countries represented in the audience (UK, Netherlands, Spain, France, Austria, Germany, Eastern Europe, Finland, and possibly others I am now forgetting).

Many of the WebCT customers were interested in the integration with Elgg. Although Elgg itself was new to most of them, the idea of a seamless workflow from WebCT to a student-centered system was appealing.

In my presentation, I had contrasted the 'formal' learning of WebCT, and the 'informal' learning environment of Elgg. One member of the audience raised an interesting point: if these types of technologies are adopted by the institution, will they become more formal, or can they retain the loose structure which makes them such a good complement to the VLE?

I also attended an ePortfolio roundtable, which was quite enlightening, but I will post on that tomorrow.

Integrated Environments

Mon, 06 Mar 2006 11:18:27 -0800

There is an interesting discussion on Bill Fitzgerald's blog regarding the integration of three systems to form a cohesive learning environment: Moodle, Drupal, and Elgg.

The sample scenarios that Bill outlines in his post could also be applied to the WebCT-Elgg integration that I have been working on.

From a WebCT-customer perspective, I would make the following comments:

many of the institutions running WebCT already have some type of portal system in place, so that infrastructure may overlap with the proposed Drupal use-case.
one big issue at institutions is faculty-support. Many institutions are driving towards one main learning management system (be that Moodle, WebCT, Blackboard, Sakai, etc), so that the technology can be scaled across campus with existing resources. A few years ago, it would not have been uncommon to see installations of WebCT, Blackboard, Course In a Box, TopClass, all on the same campus, and all being supported to some degree. As usage on campus has been increasing, I have been seeing consolidation, and a move towards fewer systems.

I am not sure that institutions would roll out both Elgg and Drupal on a large scale (especially if a campus portal is already present) - not for technical reasons, but from a faculty support perspective. While I'm certainly an advocate for choice, the key is to successful scaling is to build upon successes over time. Spreading IT or Learning Technology staff over too many products means less personal help for faculty staff getting started using these new technologies, which in turn limits the ability to deploy widely on campus.