Wednesday, April 11, 2007
My project is finished -- this stage
After many starts and stops, I have finished adding an article for every woreda -- an administrative subdivision equivalent to a district or a US county -- in Ethiopia. A small step towards balancing the Euro-American bias that Wikipedia inadvertently exhibits.
At first I thought it would be fairly simple: arrange the data so I could feed it into a perl script that would do most of the writing, then after a little proof-reading upload the product. Maybe with a bot.
What I found was that finishing the output of my script would take a lot more work than I foresaw: properly arranging the neighbors, dealing with name collisions (I discovered just how many towns, woredas and other groups shared the same name -- often people, places or things unrelated with Ethiopia), and other organizational problems.
Because my script only delivered a couple of terse paragraphs, I searched for more information to fill out these articles. Most of the information that was easiest to find concerned natural disasters -- floods and famines. Despite Ethiopia's poverty, I knew that there are many other details that could be added before I turned to these; and how would an Ethiopian feel if the article about his home was mostly about one of the devastating famines of the 20th century?
One source was to study a map of Ethiopia I purchased on Amazon, studying the landforms: I made the discovery that many of the woreda boundaries were defined by rivers. While this might be considered original research, still it was an obvious and simple deduction. Besides, if my observations were correct the documentation to confirm them would soon be coming; if they were not, then proof would be found, my guesses replaced with verifiable facts -- and the end result would be the same.
Another source I found was that the Oromia Regional government had published useful survey of her woredas -- however, slightly less than half of the 180 woredas were covered. I'm not sure whether the omission was due to a failure to complete the survey -- or due to a glitch on the Oromia State website; I could not find contact information for the website. Further, the information was not presented consistantly. Some articles provided a wealth of detail about their woredas, some presented what they had in an idiosyncratic way -- and some were clearly missing chunks of text. And there was also the problem of figuring out how to present the information I had found.
Now after all of the work, the questionable decisions, and the ebb and flow of my enthusiasm -- it's finished. And it's something that only exists on Wikipedia.
And now to write the articles on Ethiopia's towns. I estimate that I have a little more than 800 left to do.
Geoff
Update: After working through the list I created from data I used from the Ethiopian Central Statistical Agency (their equivalent of the Census Bureau), the number of articles is actually 857.
Technocrati tags:wikipedia
At first I thought it would be fairly simple: arrange the data so I could feed it into a perl script that would do most of the writing, then after a little proof-reading upload the product. Maybe with a bot.
What I found was that finishing the output of my script would take a lot more work than I foresaw: properly arranging the neighbors, dealing with name collisions (I discovered just how many towns, woredas and other groups shared the same name -- often people, places or things unrelated with Ethiopia), and other organizational problems.
Because my script only delivered a couple of terse paragraphs, I searched for more information to fill out these articles. Most of the information that was easiest to find concerned natural disasters -- floods and famines. Despite Ethiopia's poverty, I knew that there are many other details that could be added before I turned to these; and how would an Ethiopian feel if the article about his home was mostly about one of the devastating famines of the 20th century?
One source was to study a map of Ethiopia I purchased on Amazon, studying the landforms: I made the discovery that many of the woreda boundaries were defined by rivers. While this might be considered original research, still it was an obvious and simple deduction. Besides, if my observations were correct the documentation to confirm them would soon be coming; if they were not, then proof would be found, my guesses replaced with verifiable facts -- and the end result would be the same.
Another source I found was that the Oromia Regional government had published useful survey of her woredas -- however, slightly less than half of the 180 woredas were covered. I'm not sure whether the omission was due to a failure to complete the survey -- or due to a glitch on the Oromia State website; I could not find contact information for the website. Further, the information was not presented consistantly. Some articles provided a wealth of detail about their woredas, some presented what they had in an idiosyncratic way -- and some were clearly missing chunks of text. And there was also the problem of figuring out how to present the information I had found.
Now after all of the work, the questionable decisions, and the ebb and flow of my enthusiasm -- it's finished. And it's something that only exists on Wikipedia.
And now to write the articles on Ethiopia's towns. I estimate that I have a little more than 800 left to do.
Geoff
Update: After working through the list I created from data I used from the Ethiopian Central Statistical Agency (their equivalent of the Census Bureau), the number of articles is actually 857.
Technocrati tags:wikipedia
