0

Thomson Reuters have been awarded a patent for a "System and method for citation processing, presentation and transport and for validating references" that, in my eyes, looks very much like what BibTeX has been doing since 1985. If that were the case, my understanding is that BibTeX would constitute prior art and render the Patent invalid.

But I'm making these observations as a layperson with no background in law whatsoever. So perhaps my observation is wrong/incomplete or my conclusion is wrong/inaccurate (and I understand that only a court can give a legally binding answer to that)?

There is surely much to be said about the matter, so to formulate a clear question that can actually be answered, I'm looking at this from the perspective of a company that wants to sell a software using placeholders for citation processing in the manner that BibTeX or Endnote do (disclaimer: I do not own or work for such a company). The question, then, is: would it be legal to sell this software in the USA?

I suspect that this question may still be somewhat too broad since the answer obviously depends very much on what the software will actually do and how that compares to (a) the patent and (b) BibTeX. So, I'm offering the following two alternative specifications, in case this helps to give useful answers:

a) assume that the software uses the BibTeX system to process citations,

or

b) assume that although the software does not use the exact BibTeX system but its own way of constructing and parsing placeholders. However, the principle is still the same in that the software parses the manuscript for placeholders which are formatted according to predefined rules and which will trigger a database lookup to match corresponding title in order to then replace the placeholder with a reference to the title which is formatted according to predefined rules.

Improve this question

1 Answer 1

Reset to default
2

The document you linked to is a patent application. The subsequent patent is US8676780B2. Claims on patents are often narrower than the associated application. You seem fixated on the title of the patent. Patent titles are often fairly broad and generic. In order to know what this patent legally covers, one must review its claims. Here is the first claim.

  1. A computer-based method for processing one or more citations within a document, the method comprising:

    identifying in an electronic document an unformatted citation;

    parsing the identified unformatted citation and determining one or more citation terms;

    querying one or more citation libraries to find possible matching citations, each possible matching citation comprising at least a portion of the one or more citation terms;

    presenting for selecting a set of possible matching citations; and

    inserting a formatted citation based on a selected one of the set of possible matching citations into the document.

This claim seems specific to parsing an unformatted citation within another document. I'm not too familiar with BibTex, but I don't think it automatically parses unformatted citations. Rather I think it allows manipulation of citations already put into a specific format. You state in option b) "However, the principle is still the same in that the software parses the manuscript for placeholders which are formatted according to predefined rules". I think the patent is about processing where the citations are not formatted according to predefined rules. This would keep the user from having to manually entering citations into the database. As long as your software's process doesn't use each and every element of claim 1 (and claim 11), then your software should be clear of infringing this patent.

One thing to consider, but this patent was issued before the Alice decision. This Supreme Court decision has had broad impact on software related patents. Personally I find the claims to be perhaps a bit "abstract" in how it teaches a method for identifying and processing the unformatted citation. However, I am not a lawyer so I'm unqualified to comment on the current validity of this patent. As always, I suggest consulting with an actual patent lawyer in matters related to freedom to operate rather than relying on internet Q&A sites.

Improve this answer
4
  • Regarding "unformatted citations" in the patent: what is meant here is that the citations are not in their final citation format (e.g. Harvard style), not that they are completely unformatted (if they were, the algorithm wouldn't be able to identify them as citations that need to be formated). This is exactly what bibtex does and has always done. (In other words, when I say "according to predefined rules", these are not the formating rules (e.g. Harvard style) but the rules that the software requires to be able to recognize the citation. These rules can vary between softwares.)
    – Christoph
    Jul 19, 2018 at 10:18
  • Clarification question: when you say: "As long as your software's process doesn't use each and every element of claim 1 (and claim 11), then your software should be clear of infringing this patent" does this mean that a patent is only valid for the exact method described in the patent, not any mutations or variations thereof? If so, this seems to explain the tendency towards abstract language as abstraction would then be the only way to broaden the scope of the patent. And if this is so, then the big challenge is to understand where the Alice decision draws the boundary to "too abstract"...
    – Christoph
    Jul 19, 2018 at 10:28
  • @Christoph Regarding your first comment, I believe what the patent means by "unformatted citations" is that they are, in fact, unformatted with any identifiers in the document. Just like you reading a scientific paper. As for your second comment. What I mean is that if a claim says something like "a method employing steps a, b, c and d", and your system only uses a, b and d then you don't infringe. You need to implement each and every element of a claim to infringe on the claim.
    – Eric S ♦
    Jul 19, 2018 at 14:13
  • @Christoph The fact that claims 1 and 11 don't specifically describe the algorithm for detecting and parsing the unformatted citations is why I find the patent "too abstract".
    – Eric S ♦
    Jul 19, 2018 at 14:16

You must log in to answer this question.

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