MODS RECORD INFORMATION

The metadata records used by The Online Books Page are not as detailed as those that typically appear in a library catalog, but they give more detail that simple Dublin Core does. Representing these records in MODS as required by Aquifer may in some cases involve our script inferring MODS elements that are not fully supported in the underlying record. If you have any questions about the MODS records that result from these transformations, please contact us at the address below.

Notes on inferred fields

Some of the information required in the DLF Aquifer MODS profile is not explicitly recorded in Online Books Page records. Therefore, some of the required fields are automatically inferred based on metadata we do record. In some cases, these inferences may be incorrect, but we strive for as much accuracy as possible while still conforming to the Aquifer profile. Below are some notes on the inferences we make.

• originInfo: This element is required by Aquifer, and must include a date. If we have a publication or copyright date for an entry, we include that as a dateIssued or copyrightDate subelement. Otherwise, we will include the date of accession to The Online Books Page as a "questionable" dateOther subelement.
• language: This element is required by Aquifer for textual resources, and we always assume it to be English. The Online Books Page's scope is indeed currently limited to books intelligible to English epeakers, but there are a few primarily pictorial works in the collection that have no text, or that only have text in a non-English language. These will still be noted as "English" in our MODS OAI record. Some books also have text in other languages (e.g. a book that has original-language text accompanied by an English translation). These other languages are not currently noted in our MODS record.
• physicalDescription: Aquifer requires that this element be present, and requires the subelements internetMediaType and digitalOrigin.
  • internetMediaType:: Aquifer requires a note on the media types of the resources. We have format information for some, but not all of our listings. We will note the presence of HTML, PDF, and plain text where that is recorded. (There may be other formats available as well, but if so, we don't note it.) Where we do not have information on those specific formats, we simply say "application/octet-stream", a media type that applies to all data on the Web.
  • digitalOrigin:: Aquifer requires a note on whether a work is born digital or reformatted digital (where resources that include new digital content with digitized analog content are supposed to be labeled "born digital".) We do not keep track of whether our listings contain new digital content. So we assume "born digital" by default. However, if we have a recorded issued date before 1970, and no "commentary" note on the item, we will assume "reformatted digital";
• subject: All subjects (including name, geographic, and other subjects) are shown with all their subdivisions under a single "topic" subelement. While we do internally distinguish name subjects from other kinds of subjects, we do not type subdivisions, and it appears that the only subelement that allows untyped subdivisions is single topic slement.

Home -- About Us -- FAQ -- Get Involved! -- In Progress / Requested -- More Book Links

Books -- News -- Features -- Archives -- The Inside Story

Edited by John Mark Ockerbloom (onlinebooks@pobox.upenn.edu)
OBP copyrights and licenses