Wikisource:WikiProject Fast Proofreading

There are two main goals of WikiProject Fast Proofreading: the first, to educate Wikisource editors on methods to help them proofread works as effectively but at a faster pace; the second, to advocate for improvement of Wikisource's layout and functionality to better enable faster and more effective proofreading.

Members
Please add yourself to the list if you would like to participate.
 * User:PseudoSkull
 * User:Zoeannl
 * User:Paul Enenche

Justification
I have been experimenting throughout most of my time at Wikisource with methods of proofreading as fast as possible, combining both the human element of proofreading and the technological trickery necessary today to enhance the proofreading experience. I am happy to say that because of my efforts I am proofreading pages now at least twice as fast as I did when I started out here. I started with being able to proofread 40 pages in two hours of a novel with average-sized pages, and now I can proofread that same novel at around 45 to 50 pages every one hour. This counts transclusion of each chapter by the way.

Response to discouragement
Some may discouragingly say that learning to proofread faster rather than with more time for scrutiny and attention can lead to more proofreading errors that are overlooked. "Quality over quantity", right?

Well, that's at least partially true; the faster the proofreading, the more likely it is to miss a few errors here or there. However, I believe that with the correct tools, and with experience spotting likely OCR errors while proofreading, this inevitable issue can be mitigated by quite a lot more than you might think. And besides, with a script that I'm developing that could identify and bring to light possible typos within proofread or validated works, proofreading errors will get a lot more widespread attention in the near future.

"Quality over quantity" may be a fine proverb, but consider that if we had both quality and quantity, that would be magnitudes better than having either one on its own.

I'm certainly not here to deny the importance of accurate proofreading, correct formatting, and proper presentation of works being transcribed. I also think that works need to be scan-backed and proofread by Wikisource editors, and not just copied in large numbers from other sites—even Project Gutenberg. However, that does not detract from the argument that faster proofreading will lead to a larger collection of works in our digital library, which will be elaborated below.

The more, the better
It is incredibly important—one of the utmost priorities—that we maximize our collection of fully proofread and scan-backed works available at Wikisource as much as possible. This is not only true theoretically, but also pragmatically. The more works are available and presented well at Wikisource, the more visitors we will get. I think that if our collection was larger, we'd probably get a lot more mainstream attention than Project Gutenberg, because Wikisource, at least in theory, is fundamentally more usable by the people (in my personal opinion).

One of the main complaints I've gotten about Wikisource from readers is that they are unable to find the work they're looking for here, because our collection is currently so small. This would make sense, given that we're a fairly small community, and proofreading takes quite a lot of effort even for each individual work. However, if we account for the potential of faster proofreaders in the community, adding substantially larger amounts of works to our new texts, that particular audience of readers will become less and less alienated from Wikisource.

How to proofread faster
TBA

Staying healthy

 * Buy some
 * Drink lots of water during proofreading sessions
 * More TBA

Improving the site for better proofreading
TBA

Easily finding typos across the site
I think we should have a Python script, open-sourced and that can be used and remixed sitewide, that looks for common proofreading errors and typos in random works, tells you about them, and then enables you to check the page in question to see if what the script found was indeed a proofreading error.