For example, the DOI of "Disclosive ethics and information technology: disclosing facial recognition systems" is 10.1007/s10676-005-4583-2 but the parameter needed to reach the paper on CiteSeerX is as in https://citeseerx.ist.psu.edu/viewdoc/summary?doi= Would someone explain the difference?

  • 1
    Perhaps CiteSeer could answer?
    – Jon Custer
    Commented Aug 10, 2022 at 18:35

1 Answer 1


CiteSeerX seems to use its own identifier system, running on their own databases. By using their own system they guarantee that all indexed objects have a unique identifier. While they happen to call this identifier DOI, it appears to be unrelated to the regular DOI system. Perhaps they use the same name because "Digital Object Identifier" is still a good description of its role.

Anyway, CiteSeerX's source code is available at their GitHub. Of particular relevance is their DOIHandler.java class, which provides the following description:

Singleton class for thread-safe DOI service. This class is intended to maintain a database containing service configuration and records of the DOIs that the service has given out. Each service using this class should have a unique combination of SITE ID and DEPLOYMENT ID attributes set in it's database (see external documentation) and multiple instances of DOIHandler should never share databases.

The handle syntax is SITE_ID.DEP_ID.DOI_TYPE.BIN.REC, where all variables are integers and the definitions are as follows:

  • SITE_ID: unique site identifier
  • DEP_ID: deployment identifier for multiple servers on a single site
  • DOI_TYPE: marker to determine the type of object that this DOI references
  • BIN: top-level ID space, each bin can contain only 9999 REC
  • REC: low-level ID space, record identifiers

I don't know what external documentation this refers to.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .