This week’s update could be the start of something big. At Mendeley, we know that as you read, annotate, share, and organize research documents, your knowledge and expertise is encoded in your collection. Decisions such as what groups a paper belongs in, what tags are meaningful for a paper, and whether or not you’ve read the paper through to the end are all important signals about how important a given paper is and how it’s related to others. Our mission at Mendeley is to help you leverage this latent information to more effectively organize, share, and discover research. Today, we’ve taken an important step in this process by using tags to group related documents and groups together, and we’ve also added a wiki-like page for each tag to describe the concept the tag represents and to link to related concepts. Intrigued?
How it works
For every tag in our database, we have created a tag page, which looks like this:
On this page there’s a place for a description of the topic the tag refers to and there’s also a listing of groups, papers and other tags related to that tag. This seems simple enough, but it enables some great new functionality. In a tag summary you can link, wiki-style, from one topic to the next and or to authors related to that topic. What this provides is a very effective means of browsing by topic, instead of searching, to quickly find the most relevant papers and groups.
Teaching the Mendeley knowledge engine about research
I’ve often written about the personal benefits of using tags to organize your research. They grow with your collection(unlike folders), you can easily take your organization scheme with you to other programs, and they’re also just more expressive. In addition to the personal benefits of using tags, there are also collective benefits. Because any given document may have multiple tags and any given tag may be applied to multiple documents, how documents are tagged serves as a signal as to what papers are related and also what tags are related. We can then use this signal to help you find the most relevant papers fast. If you’re interested in cancer, you can ask Mendeley, “What are some concepts related to cancer?” and get an answer back: breast, colon, thyroid, skin, … You can then explore those tags, which are chosen because they occur most frequently in the set of papers also tagged “cancer”. This is all powered by the collective wisdom of everyone tagging documents on Mendeley and it’s a perfect example of how crowdsourcing is supposed to work.
“What’s in it for me?”
The biggest thing is easier discovery of more relevant research. You also have the opportunity to help define your topic area of interest using the tag wiki fields. Here you can share your knowledge of the history of the field, why it’s important, and let others know what newcomers to the field always tend to get wrong. This is a great place to explain those little idiosyncracies and historical accidents that always confuse people, like why “mesenchymal stem cell” is the wrong name, but most people use it anyways, or why a technique is named after someone who didn’t invent it. Check out your favorite tag and add your expertise today. If you don’t see your tag on the page, you can get to any tag by simply adding /tags/name+of+tag to the end of the mendeley.com URL.
Additional changes for this week’s update include under-the-hood fixes to the open API and groups. Also, grab the new iOS app if you haven’t already.
Mendeley Blog 
At last! I have wanted this capability for a long time.
However, it doesn’t seem to be complete. For instance, I have tagged two papers with ‘f-wave’, but they don’t show up at mendeley.com/f-wave.
Also, how do I access papers tagged with a tag that has spaces in it?
Is there some way to search multiple tags? E.g., when I search for “ideational”, there’s a couple hundred results, but I’m only interested in those papers tagged as both “ideational” AND “institutionalism”, because I’m doing a lit.rev. on ideational models within institutionalism. Is there some way to do that? It would be immensely helpful!
(actually, it would be helpful to be able to do that in my own collection as well – sometimes I want to find all the papers I tagged as “X” AND “Y”, but there isn’t really an easy way of doing that.
(OK, I just checked and it turns out I CAN select more than one tag to get only those with all the selected ones. Scratch that, then).
Is there currently an easy way of actually getting to the ‘tag page’ for a tag? At the moment I have to manually put the tag name into the url thus:
http://www.mendeley.com/tags/fish+aggregation/
to pick an example entirely at random.
Which seems an awkward way of doing things.
Am I being stupid, or missing something obvious? There doesn’t seem to be any other way.
David – To access tags with a space by typing in the URL in your browser, you should just be able to type the space, but if you’re pasting the link, you’ll need to urlencode the space, like so:http://www.mendeley.com/tags/gene%20therapy/
Dubi – Tag intersections aren’t “formally” supported, but in many cases you do get something like the results you were wanting.
Andrew – When you’re viewing a paper in the research catalog which has tag tags added it to it, you’ll see a tag display on the right under the readership statistics. Clicking on any of those tags will take you to the tag page for that paper. If you want to just go to a particular tag not shown on the page, you’ll need to type it in as a url, but we’ve considered putting a tag browser at mendeley.com/tag/
Very nice concept! Thanks. It is really nice to see who has contributed on the summary.
Are author-supplied keywords also used? Do they have a different status for the wiki?
Is it possible to add a tag from the webpage of an article or is it possible only from the desktop?
How long does it take to transfer a tag added in my library in Mendeley desktop to the webpage of the article? If I remove tags, are they also removed on the webpage?
Do you plan to extend the wiki-style links to external webpages and to articles in Mendeley?
Replacing spaces with + characters also works, and is a bit more human-readable and easier to type. http://www.mendeley.com/tags/gene+therapy/
MJ – It takes between 24 and 48 hours for tags to appear on the page in the catalog, but not all tags given to a document will appear on the catalog page on the web, but they’re filtered so that the public pages don’t end up full of tags only used by one person. You can add tags from either the desktop or the web – just click on the title at http:://mendeley.com/library/ and you’ll be able to edit tags and notes.
We’ll be making more improvements to how the tag pages and the wiki concept works.
Thanks for the answers!
Spanish is my language native. Can I write a wiki in spanish?
Yes, Daniel. The Wiki is open to all.
You’ve never explained fully the diference between tags and keywords fields. I’ve been using the tags field for my personal tags (like: “urgent”, “used in thesis”, “in german”), not for the “thematic tags” (I’ve put them in the keywords field. Do you plan wikifying it also? Is there a possibility to copy tags from one field to another?
Is there any plans to add implicit tags? For example, in my area there are decades worth of dedicated conferences for “automated planning”, and many of the papers are on mendeley, but the collection found ( http://www.mendeley.com/tags/automated+planning/ ) is only 28 papers.
My hunch is that you’ll only see papers from those outside of the field that are reading around and tagged it “automated planning”. Anybody in the field wouldn’t bother — I certainly haven’t tagged any of the hundreds of planning articles I have with “automated planning”. What about extracting default tags for a paper based on the meta-data? e.g., conference or journal name.
I’m also curious why “bad smells” is the number one associated tag with my field of research ;).
Tomasz: The difference between keywords and tags is that keywords are generally supplied by the author, while tags are supplied by the reader. I would suggest keeping it simple and putting all your tags in the tags field.
Christian: There is indeed a plan for implicit tags. The folders and groups in which you put documents serve as one such implicit tag, as do the disciplines of the readers, etc. We also extract the author-supplied keywords from documents where they exist. Future work includes the idea of automatically classifying documents based partly on their content and partly on how researchers share and annotate the document.
No idea about the “bad smells” thing, but I can ask someone to take a look, if you like.
When I started using Mendeley and imported my library from Zotero, all my Zotero tags got imported into Mendeley tag field. My Zotero tags were, in general, automatically populated from author-supplied keywords, so according to your distinction of tags and keywords it would have been better if they had ended up in Mendeley keywords field.* I can still identify all the entries that were originally imported from Zotero because I have a collection/folder which contains all these entries.
1. How can I go about deleting all the tags from these (~ 2000) entries? Ideally I don’t want to delete all my tags because I have added some to new documents I have added directly to Mendeley.
2. Is it possible to batch populate the keywords field of these entries?
*Perhaps you should give users the choice when they import as to whether tags should get mapped to Mendeley keywords or tags field, with an explanation of the distinction.
Also, out of curiosity, I intend to use tags which only really have personal meaning to me, e.g. “not yet read”, “have paper copy”, “journal club”. Will they end up contributing to your web catalog.
osm –
You needn’t do anything with the Keywords field. If you have author-supplied keywords in the tags field, it won’t hurt anything. For “personal meaning” tags, they’ll only appear on the catalog pages if a significant number of people use the exact same tag on the same paper.
@William, Thanks for the reply. Not sure what you mean by “needn’t do anything with the Keywords field”. They are all empty (for the documents that were imported from Zotero).
I actually very much like the concept of giving reader-supplied tags and author-supplied keywords separate status within my personal library (alhough perhaps you should merge them for Mendeley web), hence my desire to get rid of the tags that were originally imported from Zotero & populate the empty keywords fields.
However I do think it would be beneficial to have additional filter option within Mendeley desktop (bottom left) that treats tags and keywords the same (filter by Keywords/Tags)
I understand what you’re saying. We’ll consider this use case for the next release.