I do think taking categorization out of the homepage finding task would be awesome. That said, if there were examples listed after each category that we could Ctrl-F to find in the list, that would also be awesome. I know I'm a bit confused by some of the categories and where some companies fit in. The other option would be to have a blank space to type in a quick word or two that the development team could auto-sort into cateogories, which may also cut down on mistakes in the categorization process.
agreed. and just being able to roll through the URLs and not having to categorize would definitely speed things up
Whew! I've clarified the categories and separated out the categorization task, as well as tweaking the base pay to bonus ratio on the extraction and retrieval tasks. I'm currently testing two batches and sandbox and plan to have them up by the end of the day today at the latest (6pm my time, or about four hours from the time of this post).
First batch should show up in a few minutes for you guys. I'll hit go on the next one once I'm able to quadruple check that everything is running smoothly (as I've learned... sometimes triple checking isn't enough!). The other batch has over 44k records so, you know, prepare yourselves. You're looking for: Find the Homepage of a Nonprofit (rev. 3) Categorize a Nonprofit Organization (rev. 3) Retrieve Donor List for Organization (rev. 3) Extract Donor Information for an Organization (rev. 3) (Also, why is the grammar of the retrieval step so weird? I'll be fixing that later on for the sake of cosmetics.)
Looks good over here... let me know if there are any hiccups. If I don't see any hiccups in an hour or so I'll release California.
Is there any way you can segregate the N/A links for the classify. If you still do need those classified given they are more difficult to find information on it would be great if those could be their own batch with a higher pay rate. If they don't need to be classified you can simply save money by not needing to put those up.
I pulled the big boy temporarily because I was nervous the mistake from a few posts ago was more serious than it was - it'll be back in the morning
OCMP34 FYI, on the extracts, you can't click on the extract link, it is greyed out, so we have to manually copy and paste the link. It's a small thing, but just wanted to let you know.
How fast are you guys able to do the find a homepage ones? They are taking me about 90 seconds each making the pay not so great.
Has anyone seen any of the extractions post I haven't seen a single one to even set up a PandA. Wondering if I just missed them or if they aren't posting.
Hey ocmp34, Just a suggestion for the "find a homepage" batch. Some of the organization names have an ampersand in them, and when you click on the generated link it cuts off everything after the ampersand in the search As an example: Art & Creativity for Healing Inc. The google link leads to https://www.google.com/#q=Art & Creativity for Healing Inc. which only searches for "Art" so either someone will submit N/A (hopefully not) or just have to copy/paste the search manually. If you change all of the & to %26 within the search link it will perform the correct search https://www.google.com/#q=Art & Creativity for Healing Inc. Very minor but could save a lot of time and possible N/A submissions in the future. Thanks
Good morning. Working on an ampersand fix as first thing. I'll let you know what I'm able to cook up.
Hey guys - I see on the other thread you guys have written a script for the find a homepage task... if it works well, I recommend all of you using it. If you want my dev team to take a look at it and package it into something easier to use (although it looks pretty damn amazing), let me know! Thanks, ocmp34