~Solomon Ibn Gabriol~
Literature review is an essential part in writing a thesis. So how to extract the necessary information from the article you reviewed?
In this post, I’m going to share on how to write a good literature review, a legacy passed down to me from my great supervisor ^-^
Here are the steps and examples – exactly in his own words (maybe with one or two additional words, where appropriate :D)
- Simple statement of what the researchers did that is similar to your work.
- What were the authors aiming to do?
- Give a bit more detail which are needed for the next step.
- How is your work different?
- A gentle criticism. Not really needed, but it shows you have thought about what they are doing. For this case, it makes the point that these authors were not aiming at a natural interface in the sense that it is close to something already familiar.
eg: M. Chen et al.  built a 3D controller combining the Wii and several cameras. The aim was to develop a universal motion controller from current and cheap hardware that can perform general tasks in both 2D and 3D by simply changing its mode of operation.
eg: The authors stated that “the WIMP GUI may still be irreplaceable but 3D UIs will be superior when the interaction is taking place in a 3D spatial context”. Because of this, they suggested that the universal user interface should be something that can combine both 2D and 3D interaction.
They implemented this interface and demonstrated how it looked for a couple of tasks including 3D manipulation of virtual Lego Bricks.
eg: In their demo application for 3D manipulation and 3D navigation, a lot of cues were used to tell the users what interaction they were performing. A red overlay is used if an object is graspable or objects are engaged but not correctly aligned. If the objects are correctly aligned, a green overlay will be shown and the user can press a button that will trigger the ‘assemble’ command. For 6DoF browsing in the application, the user needs to point the controller upward until a semi-transparent green appears which indicates the desired operation has been triggered.
eg: This interface is very different from ours because of these additional and artificial cues. Also there is no report of user testing so it is not possible to see whether the universal interface is in any sense better.
eg: A universal motion controller is an interesting idea but it cannot be said to be natural. It is a bit like having a common steering system for cars, bicycles and boats. It might be a good idea but we would need to see some evidence.
Simple much? These steps had helped me to focus on what to write especially when I had to rewrite my background chapter just a few weeks before my submission date!
Chen, M., AlRegib, G., and Juang, B.-H. (2011). An integrated framework for universal motion control. In Proceedings of the 10th International Conference on Virtual Reality Continuum and Its Applications in Industry, VRCAI ’11, New York, NY, USA, 513–518. ACM.