Bibliomining FAQ

Q: Why do we need another new term? Why can't we just use "data mining in libraries"?

A: When you do data mining, you choose a tool from a selection of software programs. This selection of software programs is commonly known as a "library." Therefore, if you attempt to find research on data mining and libraries, you will find many articles on software libraries instead of articles on data mining of library data. In order to make it easier for researchers and practitioners to find works on the data mining of library data, a new term was created.

Q: Can I protect patron's personally identifiable information while still keeping decision-making information?

A: Yes. Take a look at this pre-print article: Steps before the Shred: Using a Data Warehouse to Save Critical Decision-Making Information before the Destruction of Library Records. That will explain how to match records and protect privacy.

Q. A freelance journalist recently asked:

> Bibliomining: is that mining of patron data in libraries or what?

Or what.

The goal is _not_ to get to the individual patron level. Instead, the goal is to look for patterns of behavior of either large groups of patrons, staff, or both to help the library managers make better management decisions.

It may be that it's useful to look at the broad demographic groups that a patron is in (like age range), but it would be unethical (and in many cases, impossible or illegal) to look at the individual patron level without the permission of the patron.

For example, in order to validate the policies governing the length of circulation, bibliomining techniques could be used to look for patterns in patron class (faculty/student in an academic library), type of work, length of actual circulation, and overdue status. If patterns are found that show that, say, books with CDs attached tend to be returned late, then the policy might need to be amended.

The goal is to indicate patterns, not locate individuals.