I'd like to create an automated method for adding new papers to a brain that serves as a repository for my "important" scientific articles.   I already have a system to import new journal articles with relevant keywords, journals, authors and/or categories into Endnote.  However, Endnote and most other citation libraries don't create a map to link these articles, it's hard to connect individual articles with each other.  

My goal is to create a Brain that contains links between all of these identifiers.  Here is a smaller version ( with only ~10 articles.

In my mind the benefit of the brain is the ease of transitioning between different parents, siblings and children.  But creating each link and node is very time consuming (journals, authors, keywords, read/unread, funding agency, etc).  

Any thoughts?
Liz Ann
Liz Ann,

The power of TheBrain lies in tuning it to your own needs. I.e. I second others in thinking that automatic import into TheBrain is not a good way forward. We have automatic import into other programs, like Endnote you have mentioned. This is a perfect program for reference management when you write your papers, and I use it for exactly this. However, linking ideas between articles in TheBrain, this is work you need to do by hand because that is where you tune your system, where you input the data in such a way that you will find them in the future.
The system you have presented in the small brain is way too complicated to maintain over years. I work with pdf's I label with year, journal, mainauthor (1993, JPerkinTrans1, Haines). This I link to the author's name and the field keywords. As children I have the title (a type, different color) and the main messages of the paper as individual thoughts. After having linked more than a thousand papers, it still works and I find very interesting connections. But, you have to put in the work. 
To piggy-back on Andreas' feedback, I agree that manually creating the links within your brain is part of what makes TB a "genius mechanism".

However, I fully sympathize with your desire to simplify and streamline the initial data entry process. For this purpose, I might recommend a slightly different strategy for organizing your papers using a combination of categories and tags:
  • Categories (Parent Thoughts)
    • Author
    • Journal
    • Topic (Main topic headings)
  • Tags
    • Read
    • Usefulness

For the status tag (read/unread), I'd suggest tagging everything that has not yet been read. This will allow you to click on the tag to bring up everything that still needs to be read (for easy review and selection). Then, as you read/review various articles, you can remove the tag, which will reduce to visual clutter in your brain from extraneous links or tags.

For "usefulness", someone recently suggested using asterisks as a way of ranking favorites, and I imagine the same strategy could be used in creating a series of "usefulness" tags:
  • Most Useful: *****
  • Least Useful: *

As for key words, it will be important to give careful thought to:
  • the distinction between main (big picture) topic headings and "keywords". Specifically, What criteria will determine whether something is a "category" or a "keyword"?
  • how you want to use "keywords".

For example, using keywords as a parent thought (your current set-up) is going to create a lot of visual clutter with connecting lines in the plex. Another option would be to set-up keywords as tags. However, this could quickly create even more visual clutter in the plex, in which case you might want to hide the tags entirely.

In light of this, it will be important to consider the purpose of identifying keywords in the first place:
  • Are keywords intended to serve primarily as search terms?
  • If so, could these keywords simply be added within each thought note to reduce visual clutter in the plex from links and/or tags?

This strategy would allow you to use the search function in TheBrain to bring up any/all of the relevant keywords, and it might provide all the keyword referencing you'd need.

Just a few initial thoughts. Hope this feedback will be helpful in considering how you design your Citation Forest.
Liz Ann,

I have been using TB for reference management for quite some time, and I have experimented with different ways to do it. I'm using TB9 now, because the preview of files (mostly pdfs in my case) really changed my approach. The advantage of TB is that you can organize the information in many different ways, resting assured that no matter how you do it, you can always retrieve it almost instantly with the search function. I organize my references in TB not because I want to have a way of retrieving something fast, because 99% of the time I can do that just using the search function with very few strokes (even just using the search function of the Mac OS), but because I want to explore and see how the information relates to each other and to how many hierarchies it belongs. That is something that most specialized reference managers can't do well, or at least they can't do visually. If it helps, this is a bird's eye view of how I organize my references:
- They are organized mostly by topics. Topics are thoughts of the type "Topics" (very imaginative). In other systems topics would be keywords or tags, but using "thoughts" allows me to generate a visual structure or hierarchy of topics. For example, if I have an article on fluid resuscitation in sepsis, it resides under "Work-References" -> "Sepsis" -> "fluid resuscitation" (of course, the beauty is that it can also have other parent thoughts if it relates to more than one topic)
- The source (a medical or scientific journal in most cases) is the abbreviated name of the journal as a tag. In that way, when I look at the list of the references under fluid resuscitation, I can see that this particular reference was published in the "CritCareMed" (Critical Care Medicine) journal
- The name of the thought for the reference always include and starts with the year of publication, and is the original name of the article. In medicine and science, sometimes that makes for very long thought names. I experimented with assigned shorter names in the past, but that can lead to duplications when you are dealing with thousands of references like in my case (my TB is going over 10K thoughts now). I include the year first, because if a see that a review on septic shock is 10 years old, I know that the article is not up-to-date and might be stale.
- I use link types. If an article relates to another but it's not worth to create a separate topic, I just link them. But if it's an editorial on that article, I use a special link of different color and with the label "Editorial" (again, very imaginative). If the article is a review on a published book, the link will be of a different color with the label "Review"
- I use plain thought  types for the articles in journals, but books will be a different thought type (three types actually, one for books that I have read, one for books that I own but haven't read, and one for books that I don't own but I might consider to buy).
- I use a few additional tags for special things. I have a red-colored tag for "Key-Reference", I have one for my own publications, one for "graphic" or "photo"
- I don't store in the TB the author information, except for selected authors, and when I do I will have a thought for the author, with the child thought as his/her publications.
Again , the whole idea is that when I see a reference, I can immediately see from the list of references under what topics is under, what's the year of publication, what journal, and if it's linked to other references, and if it's an important reference or not. Of course, the ability to preview the pdf in TB9 adds even more (visual) information.
Hope this helps.
Juan Gutierrez, M.D.
Thank you for posting this ICU_MD! I've found several ideas that I'm going to implement. [smile] 
  • I use categories instead of a topic thought type. I prefer the word "Topic"
  • I like your type identifiers for books. Previously I used the "interesting" tag, but I have so many that I'm interested in that it drowns everything else out when I view it. I prefer the "Read", "Owned & Unread", and "Buy" type idea.

Update: Incidently, I just changed the type to topic and found the (molecule?) icon under the medical section and used it as the thought icon. Wow, looks nice on a dark theme.
macOS 10.13
TheBrain 9.0.239
Unlike Juan, I actually really use the author's names a lot. The papers are connected to a parent with the name of the author: AUTHOR, papers.
The authors that appear on a regular basis have three children: AUTHOR, papers; AUTHOR, students; AUTHOR, science. The main thought with just the AUTHOR name, is connected to the institute as a parent. 
This setting allows me to see the environment of the author; why is he/she working together with this particular other author. The students tell me about the mind-set of the "descendants" of that specific school (group). And the science thought is populated whenever I see the author give a talk at a conference. The conference itself has a CONFERENCE NAAME, science child which allows me to quickly see who was at a particular conference. And when I see another talk by the same author, I continue jotting down notes where I have stopped after the last talk.
And I attach a lot of slides via the take-a-photo possibility of the mobile app.
I do not have specific tags or parents for journals because for me it is not important where the work was published.

Add a Website Forum to your website.