Check out the book "How to Read a Book" I found it to be a great help in reading and digesting texts.
How to read research paper, textbook, long text content?
11–20 of 89 posts
Re: How to read research paper, textbook, long text content?
#12The first thing to know is your goal. Your goal could be things such as: to glean core concepts and methodologies, to determine how a paper relates to another research work (such as your own), to find an answer to a specific question (or determine that the paper does not answer that question), to be able to implement what is in the paper in code, or to develop the material further in your own research.
Obviously, how you should read dense technical material will differ greatly depending on your goal.
For instance, if you are looking only for an answer to a specific question, you might prefer to do a sort of DFS on the paper. Read the intro material to get an impression of where in the paper the answer might lie, then go read those parts (in descending order of likelihood of containing the answer). A similar approach might work if you already know the area and are simply looking to see if the paper contains a particular result. Sometimes this process will still require some investment in understanding notation and terminology, but time spent doing this should be kept to a minimum.
On the other hand, if you intend to master the material and develop it further, then you might take a totally different approach where you prepare your own set of detailed notes on the paper (taking care to expand, in places where your own knowledge is lacking, techinical terms and concepts that are used in the paper but not explained using secondary references). This might also include working out many examples or thought experiments to help you internalize the logic and concepts. For a significant paper this process will take a long time, but it will be worthwhile if you truely want to understand the paper comprehensively and be able to build on it in a substantial way.
No matter your goal and approach, the point is that reading a paper productively is a systematic thing. Skimming technical papers lacksidasically is often a waste of time that doesn't produce value.
Re: How to read research paper, textbook, long text content?
#13I think of myself as an AI/deep learning practitioner, definitely not a researcher.
I am always on the lookout for papers that might help with something I need for work. I do a quick skim of papers and when I identify one or two that might be very useful, then I read them slowly and repeatedly. For me, the trick is finding material that is worth a lot of study time.
Re: How to read research paper, textbook, long text content?
#14Here's what I do: I start reading from the last paragraph. I can barely get myself to read anything long, even exams! Somehow, starting from the last paragraph, and sometimes even the last sentence, and going back, makes me actually read the text. This also worked with research papers (conclusion -> 1st paragraph). Maybe give it a try?
I do this too. There’s something about reading things backwards that more actively engages and focuses my mind.
It’s partly because seeing the conclusions first triggers a “why”, that I then have to go back and discover in the previous explanations. Most research papers put the final conclusions in the abstract, but each subsequent section and paragraph also ends with intermediate conclusions. Read those first, then go backwards to find out why.
It’s also partly because it forces my mind to actively rearrange the knowledge coming into it, rather than having it all neatly pre-arranged by the author.
Maybe other things too, but for whatever reason it helps me be an active, rather than passive, reader.
Re: How to read research paper, textbook, long text content?
#15https://web.stanford.edu/class/ee384m/Handouts/HowtoReadPape... Keshav knows what’s up. I tend to read the abstract, figures and captions, conclusion. Then i know if it’s worth really reading it, at which point i print it out and break out a pen. if it’s really REALLY worth reading, ipython usually comes out to play, too
+1 for How to Read a Paper by Keshav. I have been using this method for the entirety of my academic and industry career. Now that I run a Data Science team, I have all my new hires read this as part of onboarding.
I'll just note that this pairs wonderfully with the Whitesides paper on Writing a Paper (https://gmwgroup.harvard.edu/files/gmwgroup/files/895.pdf), which is also required reading for my team members.
Re: How to read research paper, textbook, long text content?
#16As a semi-pro writer, and on some of my longer comments, I've typically written something out, taken the last idea that came out as a result of that writing process, and pasted it to the top/front of the piece because that's the idea with the most time, effort and value I can add. Leading with that rewards the reader with the value up front, and then they read on for more, which includes my initial reaction and analogies that produced it.
Remember that most people who write are still sounding out ideas when they write, so the final paragraph or final chapter of a monograph is where to start with most things. Then I read them in reverse order.
If it's professionally edited, you can read the first few parts because the editor will have uncovered any buried ledes and popped them up into the front matter. In journalistic writing, the opinion ("context") they were trying to freight the piece with is at the end.
Re: How to read research paper, textbook, long text content?
#17Back when I was a TA if a student had come to me with the problem you are describing and the compensatory technique you are using I'd tell them they might have mild to moderate dyslexia.
If you're a university student you probably have access to a student support service that can assess you and advise you if that is the case. In the mean-time some people find using the [1] OpenDylsexic typeface helps and coloured overlays over text can be effective for some (Literally stick coloured acetate over what you're reading or change the font and background colours if on a screen). Other than that using various [2] notetaking methods will help with with both reading and retaining the knowledge. You're kind of already doing that with the slides BUT handwritten notes are generally more effective than notetaking on a computer.
[1]https://opendyslexic.org/ [2]https://medium.goodnotes.com/the-best-note-taking-methods-fo...
Re: How to read research paper, textbook, long text content?
#18https://web.stanford.edu/class/ee384m/Handouts/HowtoReadPape... Keshav knows what’s up. I tend to read the abstract, figures and captions, conclusion. Then i know if it’s worth really reading it, at which point i print it out and break out a pen. if it’s really REALLY worth reading, ipython usually comes out to play, too
+1 for How to Read a Paper by Keshav. I have been using this method for the entirety of my academic and industry career. Now that I run a Data Science team, I have all my new hires read this as part of onboarding. I'll just note that this pairs wonderfully with the Whitesides paper on Writing a Paper ( https://gmwgroup.harvard.edu/files/gmwgroup/files/895.pdf ), which is also required reading for my team members.
I like Kording, 2016 (doi.org/10.1101/088278) https://journals.plos.org/ploscompbiol/article?id=10.1371/jo... too, for the output side.
Re: How to read research paper, textbook, long text content?
#19When I was in grad school, I took over a month to go through a single paper [0]. I wasn't taking long breaks, I was trying to make headway almost every day. This wasn't procrastination or me being a slow reader. I don't need months to read 70 pages. I do need months to read 70 pages when each page has ideas that require filling several blackboards to understand.
Re: How to read research paper, textbook, long text content?
#20One should understand that not all processes in life take 45 minutes. Learning meaningful content is like going to the gym or playing the piano - you cannot expect to see any meaningful results in a day. You'll have to put in the work for a very long time, but you'll get better at it as you go. Analogy aside, not all text is written to be read in a 45 min session. Set your mind that reading and absorbing content from…
It should not take someone a week to read a research paper. Depending on the paper maybe a day but usually a few hours. Carefully studying it is something else and that might take a week, but not every paper needs that kind of attention and it's a valuable skill to learn skimming and differentiating between un-/valuable and un-/interesting papers.