How to use Citation Typing Ontology (CiTO) in your blog posts

One of the annoyances with bibliographies as we use them for scholarly papers is that is usually unclear why a particular paper was cited. It is often possible for readers to gather this information by looking at the citation in the context of the surrounding text, but this is very difficult to automate. A highly cited paper might contain a method that everybody uses, might be a review, or it might contain information that everybody disagrees with. David Shotton has thought a lot about this problem and has come up with CiTO, the Citation Typing Ontology:

CiTO, the Citation Typing Ontology, is an ontology for describing the nature of reference citations in scientific research articles and other scholarly works, both to other such publications and also to Web information resources, and for publishing these descriptions on the Semantic Web.

Using CiTO obviously means extra work for the author, so for widespread use it is very important that CiTO is as easy to use as possible. The first step would be to reduce the number of possible relationships to a manageable number, e.g. not more than ten (CiTO defines more than 20 relationships). Following a dinner discussion at the Beyond the PDF workshop, David Shotton kindly provided 10 popular CiTO relationships to Alex Wade from Microsoft Research and me. I made three little changes to the list: added “cites” as the default generic relationship, dropped “shares authors with”, as this can be done better with unique author identifiers, and added “disagrees with” to have at least one relationship that expresses disagreement.

In the next step I added these relationships to my Link to Link WordPress plugin, and I released the updated version (1.1) today. Using CiTO is an option that can be turned off, but the plugin makes it very easy to use CiTO relationships when inserting references into a blog post:

The CiTO relationship is stored in the rel attribute of the link that is created – currently as free-form text, but this can be changed to the cito:DisagreesWith format. This information can easily be extracted by computers, or made available in the bibliography to readers. The Reference Manager CiteULike is also supporting CiTO, but we need many more CiTO tools for authors.

This entry was posted in Conferences, Interviews, Presentations, Recipes, ResearchBlogging, Reviews, Snippets, Thoughts and tagged , , . Bookmark the permalink.

14 Responses to How to use Citation Typing Ontology (CiTO) in your blog posts

  1. “but we need many more CiTO tools for authors.”

    What tools do you have in mind? What should be the functionality?

  2. Martin Fenner says:

    Egon, I mean authoring tools. Whenever you insert a citation somewhere, you should have the option to add a CiTO relationship. This obviously includes reference managers, but could also mean online databases, etc.

  3. Pingback: Citation tools & Future of Publishing – Carl Boettiger

  4. Pingback: Linktipps der Woche: Quora, Bloggen in der Wissenschaft und Universität Regensburg zu Open Access | Wissenschaft und neue Medien

  5. Pingback: The Trouble with Bibliographies | Gobbledygook

  6. Andy Mabbett says:

    You have things the wrong way round. rel values indicate the relationship *of* the target resource *to* the resource on which they appear: “that page is [rel] of this page”

    Consider rel=”glossary” means “the page linked to is a glossary of this page” (and not “this page is a glossary of the linked page”)

    So your “reviews” should be “reviewed-by” (the page linked to is reviewed by this page”); “confirms” should be “confirmed-by”; “uses-data-from”should be “supplies-data-to”; etc.

  7. Andy Mabbett says:

    P.S. Some of these values could be simplified by using nouns, such as rel=”data-source”.

  8. Martin Fenner says:

    Andy, you are right that in the strict sense “rel” is not used properly here. The “rev” tag would be more appropriate, but its use is discouraged in the Microformats FAQ because of frequent misuse. ResearchBlogging uses rev=”review”.

    What you suggest makes sense from the meaning of the “rel” tag, but is not the intended use of CiTO (we want to describe why we cite these resources). For the same reason I would not want to rename the relationships to something like “data-source”.

  9. Andy Mabbett says:

    You are using the rel *attribute* (there is no such tag) wrongly; not merely in the strict sense; but in any sense.

    The correct response to the deprecation of the rev attribute is to reverse the direction of the “rev” relationship so that it fits rel; which is what I’ve done in my examples, which convey exactly the meaning which you seem to intend . They still, as you wish to do, ” describe why we cite these resources”.

    Your original values, on the other hand, wrongly claim that those resources cite (or whatever) our articles

  10. Pingback: 商用英文論文組織寫作 (第一部分:背景)十十(上) | 柯泰德英文論文編修訓練部落格

  11. Jan Polowinski says:

    I’m looking for instance data created using CiTO. Is it possible to publically access citation links that are typed with CiTO by your plugin? Does the plugin actually offer RDF (as export?), or does it only use the same/or similar terms as in CiTO and then export bibtex? If you or someone else knows about other instance data, please let me know!

  12. Martin Fenner says:

    Jan, my CiTO plugin doesn’t export RDFa or facilitate easy finding of links that use CiTO relationships. We have to build those tools.

  13. Pingback: Maurice Vanderfeesten » Blog Archive » PACE — Ping-back for Academic Citation Enhancements

  14. Pingback: Using CiTO in WordPress | JISC Open Citations