Literature Search

Reading papers and keeping track of it can be overwhelming as there are
  • so many papers out there
  • the internet will give you access to tons of results and even more links to related ones, and
  • most papers sound relevant in the first place
Below are my personal notes and strategies to cope with the flood of potential literature. Try to read as much as you can but do not be afraid if you have not read everything. You need a good feeling for what is relevant and what is not. Not everything you read will be relevant for inclusion into your paper/thesis.

Where to Find Papers?

  • Use key terms in google scholar
  • Look at forward citations, i.e. papers that cite a relevant paper
  • Look at backward citations, i.e. papers referenced in a relevant paper
  • Look at recent conference proceedings (for us that’s mostly likely IEEE VIS, EuroVis, PacificVis, ACM CHI).

How to Keeping Track of All the Papers?

Here is how I manage reading papers. Curious how you do it.
  • Keep one document per project. I organize all my document in google drive. For each paper I write, I have one document for the related work.
  • List all the titles of all the papers you find relevant during your research.
  • Keep the titles ordered by some means (e.g. year, first author last name, title, topic etc.)
  • Pick the most relevant to read. You don’t need to read the entire paper. Start with the abstract, then intro, related work, discussion, and go as far as you like. Take notes for each paper you read in your document under the respective title. I sometimes add two types of brief comments
    • how relevant I think the paper is, at the moment where I find it (e.g. I mark it with a star or some color), and
    • what the paper I am currently reading says about the paper I am putting down. E.g. the paper may say that the study is limited to X, or that the technique is seminal, etc. Such information helps later deciding how important the paper is
  • Include the URL for the paper into your document.
  • If you come across interesting papers referenced in the paper you’re reading or through any other means (e.g. Benjamin is sending you one), add the title into your document.
  • Once finished reading a paper, pick a new one and move on.
To me, this provides a good overview of the papers I have already read, and the ones I potentially want to read. Looking at the listed titles, I can decide which ones are the most important to read next.