CS 91.
421/544
Machine Learning and Data Mining
Fall 2007
Tips for Writing a Critical Paper Review
Basic layout: One page, single spaced, created by a word processor. Use margins no greater than 1 inch Brief summary of the paper. Include the names of the authors and the title of the paper being reviewed Estimation of the strengths and weaknesses of the contents of the paper Judgment concerning the grammar and organization of the paper Suggestions for the improvements for the paper Any relevant comments concerning the paper not mentioned yet. Note that critical review does not mean negative review. Search for flaws, but try to suggest how to fix the flaws. Be constructive. Mention what the papers value would be if it were improved. Some things to look for: Identify both strengths and weaknesses Identify the papers purpose. Is it important? Is it interesting? Does the paper present a solution to an important problem? Does the papers format suit the papers purpose? Is the paper organized well? Who is the audience? Are the research topics clear? Is the reported research suitable to the topic? Are meaningful conclusions stated? Are the conclusions correct? Does the paper have a clear organization, one that has a place for everything (leaves out nothing important), and also puts everything in one place (avoids redundancies). Is the language clear and intelligible? Does the writing flow? Is the use of figures, tables, examples appropriate? Would more or different figures, tables, and examples help the paper? Would alternate research designs provide greater insight? Has the paper exhibited good research design (internal validity, appropriate use of research methods and techniques, ) Was the data analysis appropriate? Did the authors reach reasonable conclusions? Are there alternative conclusions? What is new in this paper? Does the author mention related works (important)? How does the work covered in the paper related to the works mentioned by the author?