============================================================================ SemEval 2013 Reviews for Submission #37 ============================================================================ Title: ManTIME: Temporal expression identification and normalization in the TempEval-3 challenge Authors: Michele Filannino, Gavin Brown and Goran Nenadic ============================================================================ REVIEWER #1 ============================================================================ --------------------------------------------------------------------------- Reviewer's Scores --------------------------------------------------------------------------- Appropriateness: Yes Clarity/Readability: 4 Reproducibility: 4 Soundness / Correctness: 5 References / Meaningful Comparison: 3 Impact of Resources: 1 Recommendation: 3 Reviewer Confidence: 5 --------------------------------------------------------------------------- Comments --------------------------------------------------------------------------- I like this paper. It uses the very short page format to drive home a few key points that were experimentally found, such as WordNet's ineffectiveness, and the silver data's poor contribution to training. The paper is well written and easy to follow. The results table is laid out well and easy to interpret. My main suggestion is to remove Figure 1 entirely. The paper naturally describes the sequence of processors, and it is not hard to follow. The diagram is not needed, and that would free up space for 2-3 more paragraphs. I then suggest you write more about the Probabilistic correction module. Your text is so short there that it is not possible to replicate what you did. The language about averaging annotated tokens with the token in human-annotated data is a bit confusing. Expand this section out and maybe give an example. ============================================================================ REVIEWER #2 ============================================================================ --------------------------------------------------------------------------- Reviewer's Scores --------------------------------------------------------------------------- Appropriateness: Yes Clarity/Readability: 5 Reproducibility: 5 Soundness / Correctness: 5 References / Meaningful Comparison: 5 Impact of Resources: 4 Recommendation: 3 Reviewer Confidence: 4 --------------------------------------------------------------------------- Comments --------------------------------------------------------------------------- The structure and presentation has been appropriate. The system description is lucid. However, it is interesting that Wordnet features "negatively affect" performance. ============================================================================ REVIEWER #3 ============================================================================ --------------------------------------------------------------------------- Reviewer's Scores --------------------------------------------------------------------------- Appropriateness: Yes Clarity/Readability: 5 Reproducibility: 4 Soundness / Correctness: 5 References / Meaningful Comparison: 5 Impact of Resources: 4 Recommendation: 3 Reviewer Confidence: 4