How to read and take notes on research papers

Question

In each research paper, there are a lot of things that I want to highlight for later use such as definitions, explanations and concepts... While most of them focus on the topic of the paper, there are some relating to a different or broader topics, e.g. a paper about investigating performance of a specific system may introduce different benchmarks and metrics for performance evaluation and explain why those approaches are applicable for this specific situation.

Normally, I just highlight all of them, put some notes directly into the paper or using Evernote. However, when I want to look for all highlights and notes about one specific topic, I find it difficult as they scatter in different papers and documents. So, are there any tools or techniques to affectively highlighting important points and group them by topic while reading research paper?

Answer 1

This probably isn't an ideal solution; but I still find it to be the most effective in the long run:

I simply keep a log of what I read, and I specifically do this using tex. That way, you will have a searchable document for any words that you may be looking for in your notes. You can also use the makeidx package to add an index to your text document, which lets you "tag" sections of your documents with various keywords. I also make sure to cite what I read, even if it is for my own purposes. This is especially useful when you are writing a paper, since most of your references will be ready in bibtex format.

It took me a while to get used to, but I find it very nice to have a well organized research log.

Since requested in the comments, given below is what the log looks like. I don't want it to become a full latex document, but it should be sufficient enough to give you an idea. Note the index markers, which are basically "tags" that don't show up in the text.

\section{Paper #1 Name, Authors, Date, \cite{...}}
My summary of the motivation and findings of the paper, or whatever I find interesting/useful.
May be as short as a few sentences or as long as a page, depending on how relevant it is. 
\index{an important word}

\section{Paper #2 Name, Authors, Date, \cite{...}}
Same thing here.. \index{another important word or two}

\section{Paper #3 Name, Authors, Date, \cite{...}}
and so on..

\printindex

Then you can look up your page numbers in the index, which is included at the end of the document. For more information about the package, see CTAN and Wiki.

Answer 2

I currently use the software called Papers. You can use tags to arrange PDFs according to topic. Another alternative is to use "collections" since a single PDF can be placed in more than one collection.

Next, filter PDFs that match a certain "tag"/"keyword". In the example below, it's "hydrogen embrittlement":

Lastly, for each PDF, you can get a list of all highlighted text (summarised in the right column) as follows:

The page numbers in the right column are clickable, so its easy to jump to the relevant place in the article.