The Kamusi Project is seeking a student intern with an interest in business AND African languages to help develop and implement a financial sustainability plan.

The intern will work online from his/her own computer (no travel involved), in constant internet communication with other members of the Kamusi Project team in Africa, Europe, and the US.

The position is unpaid, and hours are negotiable. Rewards will include letters of reference, solid experience for the résumé, some cool Kamusi swag, and exciting and satisfying work on the cutting edge of African language technology development. (The position might transition to a
paid part-time job during the next academic year.)

On the old Kamusi site we had a fun little slideshow feature. For the original slideshow, we uploaded a bunch of pictures in a folder on our server. When someone visited the homepage, one of the pictures would load randomly, and then another random image would load 15 seconds later. The feature did not work well on all browsers, so we eventually replaced it with a script that loaded a random image from the folder, and then enabled the user to scroll through the remaining images.

When we moved to our new server, we decided that we wanted a customized slideshow that could show the pictures that Kamusi community members have uploaded to illustrate dictionary entries. Each photo would link to the entry with which it was associated, so the feature would be a good way to introduce people to new words. We also want to encourage people to use the photo uploader feature, and we're hopeful that people will be motivated to contribute images by seeing their pictures in the slideshow.

Dictionary image submitted by a Kamusi community member (click photo to see the word it illustrates):

A couple of provocative posts (Post 1 and Post 2) by Don Osborn have revived some thoughts I've been having about African languages and information technology. Don's posts have arrived at a good moment - the power here in Accra just went out for the eighth time today, but I've got about enough battery life to type some preliminary notes.

Don writes about the "long tail" and the economics of language. In sum (and without network, I can't scroll back through his posts), he is thinking about the question of "smaller" languages and how people who are born speaking them relate to their mother tongues from an economic perspective. Do people who speak these "long tail" languages feel the need to learn languages higher up the tail for economic purposes? Will such people abandon their mother tongues over time, in preference for a language that allows them to communicate with more people and therefore have expanded economic and social potential?

These are questions that I have long pondered in the context of Swahili in East Africa. Swahili is very high up the tail. With somewhere in the neighborhood of 100 million speakers throughout East and Central Africa, it is spoken by about as many people worldwide as German, roughly one out of 60 people on Planet Earth. Many of the languages of East Africa are endangered, and others are diminishing, with Swahili being the language of choice of many of those people who are drifting away from their mother tongues. I have witnessed rural households where the children are discouraged from speaking their ethnic languages in favor of Swahili, and thus cannot speak with their own grandparents in a common language (though they communicate fine, each speaking their preferred language but understanding what is said in the other).

Abdel-Karim writes from Egypt, "Regarding the content, is this [PALDO] planned to be a wikipedia-style dictionary (i.e. Entries are generated by users and some sort of community ranking/approval mechanism is in place with an optional discussion page for each entry?)."

The answer: Not exactly. There will be an editor for each language, or a team, who will be responsible for approving each entry. Remote users will be able to submit entries or edit existing entries, but all of their work must go through the official editor for that language before going live.

Before implementing the editorial tools for each language in PALDO, we need to build the data models for each component language. Languages differ in more than words, so the database that contains information about multiple languages will have to account for those differences.

We have a long task list ahead of us as we transition to our multilingual model. We've prepared a master list of what we need to accomplish, and divided the tasks into some logical categories. It's a lot to get done in a short time - the core programming needs to be done by July 1 - so if you are able to volunteer for a coding task, we'll be happy for the help.

The document listing our coding priorities is really meant for internal use, but if you are interested in seeing what we're up to, click here to look at the working draft. (This is a Google Spreadsheet, so you'll need a free Google account to view it.)

Comments, of course, are welcome on this blog.

Last night we removed the "Transition Status" section from the Kamusi Project homepage. When we moved the site from the university where the project started out, we had to re-wire a great many features. Now we're declaring that job complete.

Much of the computer programming for the Kamusi Project has been conducted by Suuch Solutions for the past few years. Suuch, a firm in Accra, Ghana, was founded by Paa Kwesi Imbeah, who originally joined the project as an in-house student programmer when we were based at Yale, along with his brother Jojoo.

The most useful book I read before starting graduate school was How to Read a Book by Mortimer J. Adler. Although I'd been a voracious reader since about age 5, "How to Read a Book" taught me the skills to efficiently get the most out of the mountains of reading material that lay before me.

Similarly, using a dictionary involves a number of skills that you might not think about when you just want to look up a word. I've observed that a lot of Kamusi Project users expect the Kamusi to magically read their minds, without pausing to think about what is necessary to get the most out of the dictionary. We try to learn from our users (for example, we developed our parser technology when we realized people were searching for entire conjugated Swahili verbs instead of looking up root forms), but there's a limit to what we can do. We cannot handle misspelled words, for example. And we do not yet have the technology to translate whole sentences.

05:10:51 Martin Benjamin: Welcome to the PALDO planning session
05:18:22 Martin Benjamin: paa kwesi is welcoming us
05:18:28 Martin Benjamin: we are about to do the agenda
05:22:37 Martin Benjamin: introductions: paa kwesi from kasahorow, currently living in london
05:23:47 Martin Benjamin: henry addo from kasahorow, living here in accra
05:26:37 Martin Benjamin: arthur buliva is with us from nairobi
05:26:44 Martin Benjamin: sampson is with us from nigeria
05:26:47 henryaddo: Paa Kwesi introducing what PALDO and kasahorow is about to a Journalist

Thank you for visiting the page for the PALDO planning meeting. The meeting is now over, and we are starting on the work of programming. However, we still want your input. Please read through the transcripts and send your thoughts to paldo at kamusiproject dot org, or add comments to this blog.

The next step is our coding marathon, happening until April 24 in Accra. If you would like to add your programming skills to the project, please join our CodeAfrica team and we'll find work for you!