PEER REVIEW FORM
Modified by Charles Elkan for use in CSE 250A, Winter 2001.  Based on the ICML'96 review form.
 

TITLE
For identification, give the title of the paper reviewed here.
 

SUMMARY
In approximately 200 words, summarize what you think are the main contributions and findings of the paper.  Avoid jargon, but use technical language as necessary.  Avoid any hint of plagiarism and write the summary strictly in your own words.  Do not paraphrase text from the paper, and do not reuse phrases from the paper.
 

IMPORTANCE
Your comments on the significance of this paper and of its claimed contributions.  How difficult is the problem attacked?  How conclusive are the findings of the paper?  Does the paper offer useful lessons of general interest?  If possible, explain specific ways in which the significance of the work described could be increased.
 

ORIGINALITY
To what extent is the reported work new?  Has any part of this paper been published before?  Are the ideas, methods, and/or findings new?  Is relevant previous research discussed adequately? Use INSPEC, Citeseer, and/or Cora to check for similar earlier work that the authors may not have cited.  If necessary, give citations for one to three papers that the authors should have cited, but did not.
 

SOUNDNESS
Are the claims of the paper established firmly?  Is the paper sound theoretically and/or experimentally?  Does the paper evaluate the strengths and limitations of the work described?  If necessary, identify specific holes in the work described.
 

CLARITY
Your comments on completeness, organization, and writing.  Does the paper describe its methods and experiments in sufficient detail for someone to replicate the work?  If necessary, make specific suggestions for improving the paper.
 

OVERALL RATING

        1       -- Absolutely unacceptable, incorrect, etc.
        2       -- Possibly correct, but totally uninteresting.
        3       -- Weak, far below quality of other papers published in top conference proceedings
        4       -- Paper has some merits, but doesn't meet top conference standards.
        5       -- Neutral score, paper on the threshold.
        6       -- Acceptable paper, above threshold for publication.
        7       -- Typical good paper.
        8       -- Very good paper, among the 20 % best at top conferences.
        9       -- Outstanding paper, a contender for best paper prize.
        10     -- Breakthrough, best paper of the year.
 

REVIEWER'S NAME