Tuesday, November 22, 2011

Wiki Suggests Articles for you to edit: just like Facebook :-)


Mmmmmm! Nice. This is what it means to persuade users into editing. Give them a taste of what they can do, based on what they have already been doing. This way, they'll never be bored. Such things are of great importance to local language wikipedias.

Filed for future reference now!

For your information.

SuggestBot is a program that attempts to help Wikipedia users find pages to edit based on their past contributions. If you want to try it out and get some personalized recommendations, please leave your user name on our request list. You can also sign up to receive suggestions regularly. If you're wondering what the bot's recommendations looks like, you can find an example on Nettrom's user talk page.

SuggestBot uses a variety of algorithms, including standard information retrieval and collaborative filtering techniques, to make suggestions. Most of it runs at the GroupLens Research Lab on various machines, some parts also run on the Wikimedia Toolserver. The recommendations are done based on downloading a list of a user's contributions (to avoid recommending articles they've recently edited). The bot was originally developed by ForteTuba.

Monday, November 21, 2011

Confusion hi confusion hai, solution ka pata nahi

This comes in as a thought from me after trying to start a new article.

What did I try?

All I wanted to do was look up an online Gujarati newspaper, find entities of interest for the Gujarati Wikipedia and start articles on them.

What is the problem?

The problem is, even if I start, how do I ensure that the article goes through a well-defined incubation process? For example, I found that there is no article on Nifti and Euro zone on the Wikipedia. So, what do I do? Well, the simplest Thing to do would be to start an article about them and write as much as I can.

However, that itself is the problem. Writing an article in vacuum, with no goals defined and no collaborators in sight is a very tough thing to do.

Forgetting the motivation part, and concentrating on the workflow process, until and unless the article is categorized into several categories, all of which are syndicated to different contributors who might take up an article and pledge to develop it; every contribution made to Wikipedia will be static.

Rethinking collaboration for Indic Languages

This calls for an overhaul of how we approach the article writing process. With English, it is different. There are a lot of English speaking editors who are also well-versed with the Internet and have an expertise in the subjects. Hence, whatever you start off with is sure to be picked up by somebody else. The numbers work like magic for them.

The numbers, on the other hand, work like a curse for us. A curse which reduces our own expectations and vision of what our local language wikipedias are capable of doing.

Not enough contributors means that unless and until a user adopts an article from the beginning to the end, an article that was submitted will never be developed by the entire community. And this is disheartening, isn't it? Wikipedia is supposed to be a collaborative effort, not an individual one. Individual efforts have to be built upon : a system/environment/ecosystem in which this doesn't happen will never develop.

Evolution of an article

In the ideal case scenario, if I start an article on, let us say "Entrpreneurship Development Institute", it should be adopted by several other contributors. this could be done through tagging the article with relevant categories - entrepreneurship, institutions, incubators, articles with new images, articles no differences - and so on. People watching these different categories could see that there is a new entry in each one of those categories and take up what the issue with them is.

After this, the development of the article could be facilitated by having a roadmap and possible developments listed out on the talk page itself.

So what now?

The process of starting an article on smaller Wikipedias has to be taught. It needs to have enough thought put into it right at the beginning so that whenever more people join and who can adopt that article, they will know that this article exists. And at the same time, it is important to let them know what the can do with the article.

Case in point: I started the article on entrepreneurship development Institute with one line. Then Sushant suggested to me that I should expand with minimal information on its origin, mission and vision and routine activities. Is there a possibility that before starting article, a new editor is taken through a wizard To help him develop a very good skeleton of the article?

What are the takeaways?

If you have read this, one of the most important takeaways for you is that the development of the English Wikipedia cannot be taken as a guarantee for the development of Indic Wikipedias in some distant future. The dynamics of English users, their motivation and the corresponding factors in the Indian counterparts, especially the local languages, are different.

The second take away is that more thought has to be put in when starting an article, making it easier for subsequent editors to adopt it.

The third takeaway is that the use of categories and watch lists has to be advocated, that is the only way by which we can ensure Ongoing evolution of submitted articles.

It would be a very pleasant sight to see a one line article submitted today grew into a 10 page treatise on the topic. it would be 1000 times more exciting if the edit history of the articles showed contribution from Dozens of users across hundreds of edits.