Keyword searches go beyond indexed data to comb all printed text in Ancestry.com’s enormous collections of city directories, school yearbooks, other books and newspapers. They scan printed pages using optical character recognition (OCR) technology, a boon to those who might not otherwise come across an ancestor’s name or hometown in these resources.
Digitized books are scanned from the cover page to final page, and every page (even a table of contents or index) is available for keyword searching, says Ancestry.com corporate genealogist Crista Cowan. In addition, Ancestry.com used an OCR process to create searchable indexes for the city directories and school yearbooks collections.
- Make it easier to spot search results found with OCR by selecting a specific collection (yearbooks, for example) from the card catalog. Perform your keyword searches within just that collection.
- In city directories, try addresses or business names as keyword searches. Search terms must be exact: West 25th St is different from W 25. Experiment with variations and browse directories from the time and place of interest to see how addresses are written. Try searching for an address without a name, with a city in the “lived in” field to narrow results.
- Yearbooks are indexed for name and estimated birth year, based on the date of the book and the person’s school class. Keep in mind the estimated birth year could be wrong for a student who entered school early or late, or for a teacher.
- Search yearbooks to see if two people (such as future spouses or friends) attended school together. Put one surname in the Surname field and the other in the keyword field.
- Be aware that some kinds of text may not be searchable because OCR software can’t “read” it, such as extra-large type, fancy fonts or sideways printing (often seen in advertisements).
- In newspapers, try keyword searches for the names of small towns. Sometimes a smaller village or municipality will have its own section in a newspaper that covers a larger area.
- Given the limitations of OCR’s accuracy, these strategies work only sometimes. In other words, when a resource relies on OCR searching, not finding someone listed isn’t proof that the person isn’t there. Browse the images yourself to be sure.