A general idea how to organize articles, to connect them within a network

Dear Zotero team,

Spending some time preparing a review, I would take a minute to propose this thought about a way to organize a library, that I think quite interesting.

The life of an article in a library does not only concern a child note linked to it, but mainly the articles which are related to it. By that, I mainly think about the previous articles cited in the reference, and the future articles citing it. As often, the importance of a reference appears by the importance of citation in other journals. I think it would be a major evolution in citation management to be able to organize references in a virtual network between them, taking into account their different types of connexion (in and out), keywords which connect them... Until now, When I want to see which articles are interesting for my research, I've to rely manually on these items.
Networking appearing everywhere (fac~bo~k...), it looks like a natural evolution for a reference manager to help using the library with network mapping. I would loved to illustrate that with a quick schematic representation, but I'm sure you know what I mean.

Do you think there is some possibility to apply this concept of networking to Zotero ?

All the best,
A.

PS : if main of my posts concern some requests, and generally not a direct thanks for the solution you offer to them, I would still renew my gratitude for the tremendous job you're doing to ease ours.
  • If you have a little programming experience you can Have a look at my zotero browser ( http://github.com/singingfish/zotero-browser ) to see one way of dealing with related items outside the main zotero interface. I used them in my annotated bibliography report which is part of that distribution.

    However the main problem with related items as they currently stand is that they're bidirectional - there's no way to store the direction of the relationship (i.e. citer or citee). However you can sort of hack around that, as if the year published is greater than the original item, it's a forward relationship, if the year is less than, it's backward. However this doesn't tell you who is citing whom, which may be a problem for your use case.

    I have heard there are plans to rework related items at some point so the links are not just the bidirectional links which they are at present.
  • Sorry but I guess my skills are limited to chemistry and reading... Nevertheless...

    The bidirectional link is indeed a trouble to get a smart connexion between articles. If you imagine mapping your references together, you will have black lines between the multiple dots, and it will be difficult to get an idea which articles are more important from the others.

    Implementing two directions to relate articles will now give you black arrows between them, which already start to give some 'life' (literally in fact, as your library is constantly evolving) and can make things much clearer.

    Now, pushing this idea by adding keywords to these relationships, you will be able to directly see your network as interdigitated subgroups. For example, applied to technical science, I would be able, connecting the references with specific links 'applications', 'concepts', 'Experimental protocols', 'measurements' (...), to colour my network map with blue, green, red, and yellow arrows. (I promise I won't complain about the choice of colors)

    Although this idea can seems a bit crazy, and is proposed here in 'feature requests', I am less asking a solution for the next weeks (although it would be nice... ;) ), than trying to see what a reference manager would be able to bring to researchers to ease our work in a future version. I'm so please to have chronologies and bibliographic reports created directly in Firefox, I can just imagine there is so much more that Zotero can bring with this.
  • Chemistry is not that different to programming ;) . Especially with the documentation and examples that I've put together.

    (speaking as a failure at industrial chemist that is).
Sign In or Register to comment.