Zotero and Ontology Integration - Scaling?

bclatran

Hi everyone!

I just signed up to use Zotero as a bibliography management software, and my research team is working to create a knowledge management ontology. I was hoping to know if anyone has been successful in integrating an ontology or a controlled vocabulary into their Zotero libraries?

I've also been trying to find more information about the dimensions of Zotero's scalability. Is there a limit, are there any roadblocks towards growing one's research libraries? What have you found out, and what would you recommend? It would be greatly appreciated if there is any insight you can share, and thank you for any help!

adamsmith

I have nothing on the ontologies question, although I believe you're not the first person to ask about this.

While there's no theoretical limit on library size, Zotero will slow down on very large libraries. The exact size of a library that starts to significantly affect usability varies by computer, library, and tolerance for lags, but is somewhere between 100,000 and 500,000 total items.

DWL-SDCA

While an ontology and a thesaurus are cousins I can help you with setting up an thesaurus and database that is fed by Zotero. Depending upon how sophisticated your information systems staff are this may be essentially free of any external cost or, if you do not have access to people with programming skills in AJAX, MySQL, PHP, Sphinx, etc., prohibitively costly. I don't have a recent version of the software that drives the system that has been stripped of my own specific named structures & functions otherwise I could offer (and have offered) this free to noncommercial entities. The last time I shared the SafetyLit database software was >10 years ago when three entities pooled their resources to pay for stripping all SafetyLit-specific things from the software but it has changed substantially since then. See: http://www.safetylit.org for what the website software can do for you. My own contact information is available on the website.

My 700,000 record online database is about 85% populated by records that have been edited in Zotero and imported from Zotero MODS files. Even complex multiple-item Boolean queries of the database require no more than one or two seconds to produce the list of results.

The thesaurus software that drives this query system originated from the open access TemaTres (http://www.vocabularyserver.com/). It allows multilingual, multi-hierarchical term patterns and extensive synonym and relationships with disambiguation features. You can obtain the software for your own server (as I did to integrate it with my database) or use their very reasonably-priced subscription server. TemaTres technical support is good if you are fluent in Portuguese but my web developers are not fluent and had no problem modifying and expanding the program without developer help. They described the thesaurus software as elegant.

vit_gri

Hi, I'm recently trying to integrate some biomedical ontologies in Zotero, I'm currently using tagging and notes

laurent.duroux

Hello!
Searching the forum space with the keywords "tag" and "ontology" brought me here. I can see that there are at least 2 recent threads touching upon a similar topic: "Semantic Relations" and "Item relation type".
I have been working with Zotero for sometimes now, and it is a great Soft!
Like others, though, I now have this need to build-in as much "controlled vocabulary" as possible. Like others, I use tags for this matter, tags which are pulled from diverse ontologies (MesH terms, Ontobee, BioPortal, etc...), and input manually, for the time being. On this specific matter comes 2 questions:
1. For any entry, in the Tags editor field, would it be possible to create tag groups, to keep track of the source of tags, maybe differentiate them with colour-codes like in the Tag capture field (bottom left). This goes with the idea to associate an entry with distinct ontology-specific vocabularies, a first necessary step to integrate an entry into an ontology.
2. Would it be possible to populate automatically the tag field of any given entry, from the output of a web-query (trough a browser interface)? I am thinking here of the MeSH on Demand tool from MeSH DB, which allows to pull a list of MeSH terms from a free text input. This free text could be from the Abstract in the Info field.

On the matter of connecting (or integrating) an Ontology editor and builder (in)to Zotero (see "item relation type" thread), this seems to be another challenge, as Tags would need to be organized into Triple Stores to formally describe the nature of their relations. At this point Zotero triple store tables could be read and handled by Protégé for example.

This is quite a project for developers, but surely one which would be very useful!
I am not a developer myself, but seriously looking into this.