How we get the Keying Wiki
Have you ever been looking at or searching a particularly intriguing collection and wondered, “How do they get the names from these images? I wonder what goes into that?” You have? Great!
In order to get that data from the images, you need instructions as to how to extract it. These instructions are provided in the form of the Wiki Instructions and is prepared by the Pre-Production team.
The most important part of the Wiki Instructions is the keying instructions, which are the actual written instructions as to how to extract the data. The goal of these instructions is to explain to you, on a very broad scale, how to find and key each field for each record. We try to avoid getting into too many small scale “if-then” explanations as this will tend to complicate the project, confuse the keyers or even create contradictions between different instructions.
The instructions are written on a field by field basis. When writing these instructions, we try to follow a specific format which focuses on three main principles or steps.
WHAT TO KEY
When instructing what to key, we often tell you whether you should key using a dictionary for assistance or whether you should just key the data as seen. This allows for variations in what the person who recorded the data years ago actually wrote. These values might need to be cleaned up or normalized later on but you will not have to try and “fit” or “match” the data with what is in the dictionary. Often the dictionary values are helpful in determining what the records actually say. An example of the instruction might be: “Key the name of the province as seen using the dictionary provided for assistance.”
WHERE DATA IS LOCATED
Probably the most important part of the keying instructions is helping you correctly identify where the different parts of a record are located on an image. While we do provide image samples with callouts, it is only a sampling. Because you will have to look at and classify each image and then key all records, it is important that you are able to find the fields in all cases based on a written instruction. When directing you to the location of the data on an image, we will often use key words that are found on a printed form such as column headers. An example of instructions from the image below might be: “The province will be found on the header of each form on the line next to the printed words ‘Province.’” Or “The surname will be found in the section titled ‘Name and Residence’ in column 1 titled ‘Name of each person in family, household or institution.'”
A large portion of the content that we publish does not come in neat columnar or on pre-printed forms. If there are not specific visual “landmarks” on the image, the written instructions will often include key words to assist you in keying. This is particularly common in foreign language content. An example might be: ‘The birth date will follow the word “Geb” or “Geboren’” or perhaps ‘Key the birth day as a numeric entry from the birth date section in the form of geboren den, or am Birth Day-Birth Month-Birth Year.’”
ANY OTHER INFORMATION THAT MIGHT BE PERTINENT TO KEYING
Once you have been instructed on what to key and where to find the particular values for that field, we try to provide you with anything else that might be of use. Oftentimes, this portion of the instructions is where differences between different form types might exist or perhaps information about what the data might look like (date order, abbreviations, etc.). We can also give special instructions on how to handle certain issues that have been discovered during the analysis. This might include things such as how an enumerator handled the same value by using a ditto mark or how the priest recorded something in the margins.
As with all instructions, there tends to be a fine line between too little and too much. As mentioned earlier, the challenge is to direct you how to key based on the organization of the majority of the content and to avoid too many smaller edge-case instructions that could cause confusion or contradict other instructions.
Luckily, you have these keying instructions, sample images, dictionaries, AWAP keying standards as well as a relationship with Anna and Sean and the community as a whole to assist you. You may still find new and interesting things that we didn’t account for but for the most part, the keying instructions should allow you to extract loads of great data in an amazingly quick manner.
– Chris Spencer
I’m new to the AWAP, but have been extracting records from Ancestry for years. I’ve done a few trees for folks I’ve met in the Alzheimer’s home my mom is living in. These folks are great and I’ve learned and shared a lot of history and geography of their ancestors along the way.
The last year or so, I’ve worked on extracting information from the Italy Civil Registration Records, 1900-1929. In particular I’ve focused on the Siracusa > Carlentini > Atti di “books”.
Is there a database set up for these records?
I have extracted about 2000 birth, death and marriage records that I have put in an Excel spreadsheet.
Is there someone I could talk to to import this data into AWAP?
thanks Jim