xixtas open

News and opinion about open content and human powered search.

Friday, November 23, 2007

Licensing Woes

There's an interesting conversation (here and here) right now about licensing on the Wikimedia foundation-l mailing list. The take-away seems to be that many people at Wikimedia (including perhaps Jimbo Wales) would like to harmonize the GFDL (which is the license that both Wikipedia and Wikibooks are released under) with the Creative Commons Share-Alike licenses. ONe suggestion for the harmonizing process would apparently involve changing the next generation GFDL to closely mirror CCSA2.5. Relying on the agreement to license content under future versions of the GFDL that users accept when they upload content to wikimedia projects.

I may have inadvertently touched off the discussion with a request to multi-license some of my books at Wikijunior. The primary reason I would like to dual license is because the open education community recognizes and seems to have a strong preference for these licenses. It also make cross-pollination with other CCSA works much easier. I'm no expert, but I still don't understand why books that I have created and uploaded to Wikibooks, cannot be dual licensed. For right now I'm taking a wait and see approach.

Thursday, November 8, 2007

Pandora vs. Last.fm

I've been playing around with Pandora recently, and I have to say that I'm impressed. When I ask it to create a radio station based on an artist it really does a good job of picking songs. Much better than I could do. I'm listening to Modest Mouse radio now, and it's played a song by the Strokes, Bishop Allen, and one by Death Cab for Cutie. The songs Pandora picked actually _sound_ like Modest Mouse. Likewise with Cat Power last night (Cranberries, Alina Simone, Damien Jurado). The backbone of Pandora is the music genome project; people who not only like music, but also know about how it's made, classifying songs and artists. In this sense it is not only human powered, but expert powered and this could make all the difference.

By comparison, last.fm seems clumsy; relying on genres, crowd-sourcing and informal tagging to group songs and artists. I've found some good music with last.fm and I've spent many hours training it. But it still misses almost as often as it hits for me.

I'm looking forward to playing with Pandora some more. Meanwhile here's a link to my first attempt at a Pandora station: Spin Sad.

Tuesday, October 30, 2007

An Open Letter to the Wikimedia Foundation about Wikijunior

Here's the text of a letter I sent to the Wikimedia Foundation mailing list about Wikijunior.


My name is Randy Wilson and I have been working at Wikijunior for the last year. Because many of you are probably not terribly familiar with Wikijunior, I would like to give you an overview of the project and also give my opinion about some of the recent conversation.

Wikijunior was launched by the Wikimedia Foundation in November of 2004. There were three original titles, Wikijunior Big Cats, Wikijunior The Solar System, and Wikijunior South America. In the three years since, Wikijunior grown a great deal. Currently there are more than twenty books in active development. The books include diverse topics including books about history, culture, language, science and mathematics. The books are illustrated with photographs and illustrations and contain interesting and valuable information appropriate for young people.

The project has adopted a formal mission statement "What is Wikijunior?" which highlights the core values of the project. The core values of Wikijunior are: kid friendliness, collaboration, fun, reliability and openness. The project also has defined a manual of style to help consistency. We have a structured method of creating new titles which is intended to encourage collaboration on new titles and continue to grow the project. We have developed an effective methodology for collectively creating books from modules and tracking the progress of each module.

Wikijunior has several books that are available in an html print version as well as a pdf download. We have at least one book that has been published as a print book and another that has been made into an ebook and a streaming video book. Several of our books have been named as "featured books" on Wikibooks.

While we have a lot to be proud of, there are a few areas where people express concern related to Wikijunior. The age-appropriateness of the books is spotty. Most of the books contain some modules (think chapters) with excellent age-appropriate content. However it is fair to say that many of the books have reading levels that are too high for the intended audience. The quality of the information is not always as good as it needs to be, there are too many factual errors and too much unsourced information to be a reliable resource to be used in schools. Some believe the new title policy restricts the growth of the project and turns off potential contributors. Still others believe that the wiki nature of Wikijunior makes it an unsuitable platform for educational content.

Unfortunately, the number of contributors and contributions has dropped to an alarming low over the past few months. Our ideal contributor would be someone with a background in children's education, or children working with adult assistance; so these are the kind of contributors we would like to reach out to. Private discussions with a few contributors have led me to believe that more primary educators would be willing to contribute and use Wikijunior except that the problems of vandalism and potentially incorrect content are difficult for them to get past. You see, here in the US a teacher could not only get fired for showing her class some of the things I've seen on a vandalized Wiki page, but could actually end up in jail. There is not a big problem with vandalism at Wikijunior and the Wikibooks admins are generally effective vandal fighters. Still, it has happened that there was a module vandalized with profanity stuck into the middle of a paragraph that stayed that way for more than four months.

The idea of creating a static site at Wikijunior.org has been discussed before. In fact there was an announcement some time ago that the board had decided to move forward and do just that. Some of that discussion in the staff lounge history at Wikibooks. Having a secondary static version would minimize the risk primary school teachers run of exposing their students to foul language, offensive content, and incorrect information. Here's one way that I think it could work effectively with a minimum of developer time:

1) Wikijunior.org would run Mediawiki software and have the import function enabled from Wikibooks.
2) Wikijunior.org would have a stripped down default skin with larger fonts and no sidebar as its default. (See for example: http://www.mono-project.com.)
3) Administrators could change their skin to make the editing tools visible.
4) All modules on Wikijunior.org would be editable by administrators only.
5) Book development would continue on Wikibooks and be open to anonymous edits as it is now.
6) When a module was ready, it would be transwikied to the Wikijunior.org site by an administrator.
7) Each module would have a "contributing" link that would link back
to the Wikijunior development site on Wikibooks.

There is another potential side benefit of a static version. I think that a static online version would be a major step toward physical publication. The problem with production for print is that everything needs to come together at the right moment. A static site would be useful for "staging" a printable version piece by piece and allowing contributors to visualize the completed version, watching it grow as each module was vetted and moved over. The Solar System is an example of a book that is really close to being ready. Primarily it needs someone to go through and fix the many layout issues, select the content to be included and give it an overall narrative.

I could see Wikijunior.org growing into an outlet for free works from other projects. As long as the work is in keeping with Wikijunior's core values and is intended to be consumed by children, I think that it would be great to have other types of content. I think that material from other GFDL projects that is suitable for kids could be incorporated into a Wikijunior.org site fairly readily. For instance a selected set from the gallery at Commons:Featured pictures/Animals/Birds might be suitable. I didn't see much on Wikiversity and Wikieducator that seemed aimed at children though but perhaps someday soon. I don't think that a site at Wikijunior.org needs to be specifically for "books and magazines". I think it could certainly host podcasts, videos and other learning-oriented open-content media (in the future.)

In relation to networking Wikijunior, we definitely need to gain more mindshare among the open education community. I don't know how to get people to blog about Wikijunior. Cross links with related projects seems like a good idea to me. I would love to talk with anyone active in such an organization, or anyone who has an open-education blog.

So far, attempts at an English-language kids encyclopedia (Wikikids) haven't been terribly successful. I think that Wikijunior could offer to host one if Wikibookians didn't object. I can see us developing both a kids encyclopedia and a kids dictionary in the future. I personally don't see a problem with keeping it all under the same umbrella. But I think this would require discussion first.

I realize this has been a very long email, but I hope that it has given some of you who are unfamiliar with the project a better understanding of what Wikijunior was, is, and could be.

Thank you,

Randy Wilson,

Tuesday, October 23, 2007

Search Engine Fatigue

According to a new, independent Kelton Research survey commissioned by Autobytel 72 percent of online searchers become impatient or frustrated when they are unable to quickly find the exact information they need. Of those three out of four physically leave their computer without finding the information they are looking for. This means fifty four percent of Internet searchers sometimes "give up" on their Internet searches without finding the information they are looking for.

The comprehensiveness of search engines can lead to information overload and frustration for users. Search can be improved, and finding better ways to help searchers find the most relevant Internet resources for their needs is the great promise of human-powered search. Three areas where human-powered search has great potential are: trustworthiness, relevance, and determining intent. All of these could help narrow search engine results turning a list of hundreds of results into a list of dozens.


I, the author of this work, hereby publish it under the following licenses:

Creative Commons License
Creative Commons Attribution iconCreative Commons Share Alike icon
This work is licensed under the Creative Commons Attribution-ShareAlike license versions 3.0, 2.5, 2.0, and 1.0.
You may select the license of your choice.