Architecture and Code Optimization (TACO)


Search TACO
enter search term and/or author name

Social Media

Information and guidelines for reviewers


Submissions to TACO are reviewed with criteria traditional to the process of selecting significant, archival articles.

I. Refereeing Papers submitted to TACO

Papers must be of high quality and fall within the scope of the journal. There are four main ingredients to an acceptable paper:

  1. High technical quality;
  2. High relevance and interest;
  3. High novelty and significant contributions;
  4. Effective presentation.

Few papers excel in all of these, but a substandard level in any of the four ingredients is sufficient ground for rejection. Many papers require substantial revisions before acceptance, and reviewers should not hesitate to recommend that a paper be rejected pending changes that are required for completeness, correctness, or to substantially improve clarity. For the reviewers, all items in the Review Form should be carefully considered.

Rarely do all reviewers agree on a submission, so the detailed evaluation of the merits and deficiencies of a paper is needed. If you like the paper immensely, do not assume other reviewers will feel the same way. Describe in detail what you think is important about it and how it will contribute to theory or practice. If you are sure the paper should be rejected, you should explain why, politely, but in detail because other reviewers may recommend acceptance. Often, first submissions receive a Revise and Resubmit recommendation; for the authors of these papers, your detailed comments will be of tremendous use in guiding the revision of their work. Even if the paper is not accepted, the work may be published elsewhere, and the authors will still benefit from your constructive suggestions.

The tone of your review is important. When you write an anonymous review, you are acting as a representative of our field and this journal. It is always possible to be constructive and firm without being hostile.

II. Instructions for Referees

  • A. One of TACO's Associate Editors will email you, asking you to review a submission.  Email the Associate Editor back, letting him/her know whether or not you will review the submission.  If you agree to complete the review, the Associate Editor will assign you to the submission in Manuscript Central.

  • B. Go to Manuscript Central ( and click on "Referee Center."  Under "Manuscripts Pending Review," click on the title to view the submission. The Associate Editor asking you to review the paper will create an ACM Manuscript account for you if one does not already exist. If you already have an account, then the paper to review should appear under your existing account. To retrieve your account name and password, at the ACM manuscript login prompt, choose the "check for existing account" and enter the email address used to send you the request for the review. This will then send to you your login and password.
  • C. Download the submission. Spend sufficient time with the submission to evaluate its overall quality, novelty, and contributions.
  • E. When you are ready to submit your review in Manuscript Central, return to your "Referee Center." Click on the "Review" button, and then the "Score Manuscript."  Complete the online reviewer form, and submit your review. Please construct your review around the questions below.  You might find it useful to construct a review in a word processor and then copy and paste it into the comments box.  Or, you may choose to upload the file itself.  You may leave the review open if you are not finished submitting your review, or submit your review to the Associate Editor.

III. Guidelines for Composing the Review

The review form contains an area for confidential comments to the Editors. All private information and questions are clearly marked. These remarks (as well as referee information) are strictly confidential and are not sent to the author.

The overall recommendation should be based on the following five criteria:

  • Relevance: How relevant is the submission to the Topics covered by TACO.
  • Originality/Novelty: How innovative is the research being presented? If the research was previously published in a conference, the submission must include 25% new material. Is the submission significantly different or better than previously 'published' material?
  • Importance: How important is this area, and the presented solutions. Were significant contributions made?
  • Technically Sound: Is the submission technically sound? Are the concepts correct and accurate? Does the submission operate correctly? Is the technique complete?
  • Clarity: Are the techniques presented clearly? Is it well written?

Please make comments that are as constructive as possible. Authors spend significant time to create and submit their paper. Your comments should be meaningful and respectful.

IV. Overall Recommendation

For the overall recommendation there are four options: Accept, Minor Revision, Major Revision, and Reject.

  • Accept - An Accept with no revision means that the submission is perfect and there are no suggestions for improvement. The paper is ready for publication.

  • Minor Revision - A Minor revision should only be used for papers that have a clear contribution, and there are only small changes that need to be made to make the paper ready for publication. A minor revision usually means that only textual changes are needed. If changes to the methodology is needed or simulations need to be run, then major revision should probably be given, unless it is felt that the authors probably already have the additional data or the results can be gathered very quickly.
  • Major Revision - Major revision means that the submission has a clear contribution and there is enough evidence in the current submission to demonstrate that contribution. Major revision should only be used if the paper excels in (1) relevance and interest to TACO's audience, and (2) the submission has significant contributions and/or is novel, and either technical quality and/or presentation needs major revision. For technical quality, this means that a few experiments or comparisons may be needed to complete the publication to make it journal quality. BUT the evaluation of the contribution of the paper should not be resting upon the outcome of these additional results. If the paper is missing important results to evaluate its contribution, then the paper should be classified as Reject. A paper classified as major revision is "conditionally accepted" based on adequately making the suggested major revisions. Therefore, there is a very high probability, but not 100%, that the paper will be accepted.
  • Reject - This rating is used when the submission is off topic for TACO, it is an incremental contribution over prior art, or a more complete submission is needed to better evaluate the ideas presented. This classification is also used by a reviewer when a paper shows that it might have a potential contribution and the topic is of interest to TACO, but not enough information is provided in the submission to determine this. Papers in this category often need more comparisons to prior work to determine if the proposed approach advances the state of the art, or needs a major rewrite to allow certain points of the paper to be understood. The authors can revise, run new experiments, and decide to potentially submit to TACO as a new submission, or to a different conference or journal at a later date.

A reviewer suggesting Major Revision or feels that the paper should be Rejected and then resubmitted to TACO needs to provide an itemized list describing exactly what the authors need to address in their revision for it to be acceptable.

All ACM Journals | See Full Journal Index