Article Reviews - Topics in Software Engineering

Every class you are required to write a brief, one to two page review of all articles assigned for that class. The syllabus and reading list describe the readings, and when they are due.

Article reviews have multiple purposes:

• Improve understanding of the readings by summarizing key points,
• Develop a critical understanding of materials by making subjective judgements about articles, and writing your subjective questions/impressions of the article,
• Improve your speed and ability at reviewing scientific articles (these are similar to peer reviews for conferences and journals)
• Demonstrate that you have read the article.

Each paper review should be typewritten and include (in this order):

• The title, and first author's name
• The main point that the article seemed to make (2-5 sentences). This should be in a separate section from the 2-3 paragraphs of discussion.
• Two subjective numerical ratings on a 1-to-6 scale (1 low, 6 high):
  • a) How important is the material covered in the article?
  • b) How well-written was the article?
  • Note that the numerical ratings do not need to have written justification.
• Two to three paragraphs concerning the content of the article, containing either:
  • a) A question about the article, such as one that you or someone reading the paper for the first time might have to stop and study, look elsewhere, or reread to find an answer. Questions should be accompanied by an elaboration of the question, and/or a discussion of its relevance.
  • b) A comment on the article, such as discussion of how you might apply the results in a different setting, classification of the system or results as compared to other research, comparison and/or evaluation of method or methods.
  • c) What you liked, disliked, found interesting or found unclear in the article (and a description of why).
  • d) Discussion of how the article relates to other articles you have read for the class.

Also, see How to Read an Engineering Research Paper by Bill Griswold for additional ideas (http://www-cse.ucsd.edu/users/wgg/CSE210/howtoread.html).

Paper evaluations will be graded according to the following scale: 0: not submitted, 1: marginal, 2: what was expected, 3: outstanding. A score of 2 receives full credit for the assignment. Grades of 3 are uncommon, and give extra credit for reviews that demonstrate exceptional insight, understanding, or analysis of a particular article.

Reviews may be considered marginal for any of the following reasons:

• Substantial grammar or spelling errors
• Not typewritten
• Review fails to demonstrate that article was fully read and understood
• Questions about the article are not accompanied by a discussion of the relevance of the question, or the question is trivial or non-germane
• Discussion of likes/dislikes/interesting/confusion in the article without an accompanying description of why you had that impression
• The review focuses on the writing in the paper, and not the content of the paper
• It is impossible to distinguish the summary of the paper from the discussion on the paper.
• A significantly below average quantity of text in the review

Since every article review clearly states the title and author of the article at the beginning of the review, it is not necessary to cite quotations from the article under review (you must quote all quotations, however). Quotations from other articles must be cited.

Last updated: 9/24/2003