Tips for Performing a Thorough Literature Search 

It is critical that you perform a thorough literature search on your research topic. This is how you will come to know what you are doing that is an improvement over what others have tried. These days it is extraordinarily easy due to extensive electronic database subscriptions through the library.  

First, perform an electronic search. If you are working in the HCI area, minimally, use the following databases with a good variety of keyword search combinations:

• IEEE Database
• ACM Database
• SpringerLink
  
• Elsevier
• Medline
• PsychINFO
• Library of Congress (for books/texts)
• Google scholar
• Basic web searches with your favorite search engine

The MIT library has a great set of links summarized in useful ways online for various types of searches. 

You will quickly find most of the key prior work. In some cases, yes, you will need to go to the library to find it. Read the bibliographic management tips to save yourself a huge amount of trouble.

Once you have found and read the papers that are most relevant to your work, you should then do the following to ensure you have the absolute latest results:

• For the key authors, find their home pages and see if they have any very recent publications. You might consider writing email to the most critical people, nicely asking if they are aware of any recent work following up on their last published paper.
• Perform a citation search, which will tell you who (if anyone) has cited the papers you believe to be the most relevant. Several of the electonic databases will give you this information.
• Search online for the schedule for the most recent major conferences where some new work might have appeared (but might not yet be in the electronic databases). For novel health technologies, I'd suggest you skip down the titles of papers most recently presented at these conferences:
  • UbiComp
  • CHI
  • Pervasive
  • Pervasive Healthcare
  • ISWC
  • UbiHealth Workshop

At this point you should be quite familiar with the relevant prior work. Expect to spend some time on this, particularly reading the papers that you do find. You should get copies of the most relevant papers to your advisor as you find them, indicating why you think they are important for your work. 

A caveat: the world is a big place with many smart people doing many interesting things. In most cases, a thorough search will turn up prior work that is more closely related to your ideas(s) than you may have expected. Don't get discouraged. Finding prior work doesn't mean that your idea isn't interesting or worth pursuing, it just means you have done your due diligence. Your job when writing up your reserach is to differentiate what you have done from the this prior work and show how you've extended the state of the art. 

Stephen Intille's Thesis Development and Writing Tips

Last updated: 9/28/2010