DS504/CS586 - Big Data Analytics - Spring 2020

Reviews/Critiques:

What is a critique.

The critique should not be a mere summary of the material. Rather, a critique is typically a short written description of the key contributions and weaknesses of a paper, and why you think so. A critique should address several or possibly all of the following:

1. what is the one most important point of the paper?
2. arguments for why the work is notable or novel or neither,
3. if the problem the paper tackles is important and or not,
4. if the proposed solution is potentially useful or not,
5. are the assumptions clearly specified and are they reasonable and practically valid,
6. point out additional contexts where the same idea or technology could be applied, relate the work to another paper that you find during your literature search, or that we have read in class up to that point, if any,
7. have the proposed ideas been evaluated in some form, if so how, and how thorough is that evaluation,
8. identify a list of possible future research tasks to make the proposed work even better, develop a different solution strategy, or to drop some of the given assumptions, and so on.

In general, I would advise you to pose yourself concrete questions (in fact, state those explicitly in your written document) that you then proceed to answer in your critique related to the reading in order to help you organize your thoughts and thus writing. The hardest part of writing a good critique is probably for you to figure out what are good questions to ask yourself about the reading, then to explicitly formulate those questions. Once that is accomplished, you can proceed to answer them one by one. Examples of such questions may be : (a) what are the main features of applications that these proposed techniques service, (b) what are hidden or stated assumptions about the applications that the paper makes, (c) what are key differences between this new result and the prior technology, and so on.