I meant the entire paper, and not re-translation, but "translation" with no ability to read source text, consider it blacked out. The names/numbers/citations greatly constrain meaning. This could be done with any text by blacking out all of the text except for some domain-specific words and citations, and see how close someone familiar with the domain could come to the original by filling in the blanks. Just an exercise in constrained writing for fun!

Mike Linksvayer at 2014-04-23T15:28:10Z

I see what you mean now. In the article we wrote it can be easier as we  opted to keep the original English key terms (eg. titles of the ODs) as well as their Chinese translation. They serve as a miniature look-up table for domain-specific terms. I guess if we dropped the English key terms in our text (so you get to see our translations of Access but not the term "Access" itself), it can be more interesting. I hope by now I have not confused neither you nor myself. :-)

Tyng-Ruey Chuang at 2014-04-24T06:04:44Z

I saw that the text is now also at http://lucien.cc/?p=5939 so I read an autotranslation. Looks like a good summary of open government data stuff, glad to see Yu and Robinson prominent. I have one critique, which is my usual for this topic: FLOSS mentioned as a forerunner, but not as directly related or even necessary for realizing promise of open government data. Instead, something is said about Socrata and Microsoft positioning their proprietary software for open government data. Hmm, where can I find one of my mini-rants on this ... hmm, I thought I had something more direct, but http://gondwanaland.com/mlog/2012/10/07/open-data-nuance/ for now. I think it is a huge mistake for "open data" (government or otherwise) to treat the software generating/hosting/querying/analyzing/verifying data as neutral, ie OK to be non-auditable and proprietary.

Mike Linksvayer at 2014-04-25T19:19:55Z

Tyng-Ruey Chuang likes this.

Mike, your critique is certainly valid and I am glad you raise it. There is no excuse in not citing the use of CKAN etc. in hosting/using (open) datasets by government agencies and users. I think what was slipped in was a desire to say something to the effect that "the other" are also playing up the open government data hubba hubba with their products and services; we failed to state it in a proper context.

I also wish I can say more in the article about the process from which the data is produced and released. If the process is not open and transparent, ones cannot be sure if there exist systematic biases or "bugs" in the data generation process, and for them to be fixed. The parallel to open source in open data indeed is manyfold.

Tyng-Ruey Chuang at 2014-04-26T06:49:53Z

Mike Linksvayer likes this.