Before discussing options for exposing your content to Zotero, ask yourself “Do I want users to capture data about individual pages of my site, or do I want them to capture information about one or more resources discussed on that page?”

Let's clarify this with two examples. Consider an article in an online newspaper and a bibliographic record in a library catalog. In the first case you will want to provide information about the page the user is looking at: information about the article. In contrast, the library catalog page of the second case is in itself uninteresting. Here, the user is interested in information about the resources described on that page.

Many of the strategies discussed below can fit either case, but it is important that you identify exactly what you want Zotero to extract from your site.

• Use a Zotero-ready web application: Several web applications work with Zotero out-of-the-box, or with minimal configuration or plugins. If one of these platforms suits your needs this will be the quickest and easiest option, requiring minimal technical proficiency. These applications include the content management systems Omeka and WordPress, and the web-based bibliographic managers refbase and Bebop BibTeX publisher.
  Most libraries, archives, and museums use one of Zotero-compatible OPAC (Online Public Access Catalog) software packages listed below to provide online access to their collections. In some cases Zotero translators have to be tweaked to work when institutions have modified the software.
  • SIRSI
  • Aleph
  • DRA
  • Dynix
  • Evergreen
  • GEAC
  • InnoPAC
  • Koha
  • TLC/YouSeeMore
  • Voyager (WebVoyage)
  • VTLS
• Expose your meta data using an open standard: Alternatively, you can allow Zotero (and other clients) to use your information by using one or more of the open standards listed below:
  • Embedded RDF
  • COinS: According to the COinS website, “COinS (ContextObjects in Spans) is a simple, ad hoc community specification for publishing OpenURL references in HTML.” If the somewhat limited and inflexible categories of information available in OpenURL are suitable for your needs, using COinS is a relatively easy and lightweight means of making your data available to Zotero. For dynamic sites, see COinS in PHP, and for static pages, see the COinS Generator.
  • Dublin Core XML
  • MARC
  • unAPI: As described on the unAPI website, “unAPI is a tiny HTTP API for the few basic operations necessary to copy discrete, identified content from any kind of web application.” For our purposes, unAPI allows you to serve up bibliographic information in a variety of different bibliographic formats for Zotero to automatically ingest. MODS XML, developed by the Library of Congress, is probably the richest standard format available. The older, flat MARC format is also supported, and can be used if your application already generates it. Zotero/Dublin-Core-like RDF is also supported; Zotero is able to import the most information from Zotero-generated RDF, but the ontology is still subject to change and it has not yet been adopted by others. While other “legacy” flat file formats (BibTeX, RIS, and Bibix/EndNote(R)/Refer) offer a much more limited import, code library support is very good and they are simpler to generate.
• Create a Zotero translator: Zotero support can also be added with a Zotero translator (with the disadvantage that this solution is limited to Zotero). Translators are written in JavaScript. If you wish to distribute a translator to all Zotero users, post to the zotero-dev mailing list.