This technique report introduces TexSmart, a NLU system that support NER/FGNER and clustering. TexSmart is built on top of clustering of tokens. They first mined many is-a relations from the web and cluster them into thousands of categories. These categories are manually given a hierarchical label. During testing, a mention and its context is taken as input to compute the similarity against each cluster to predict its fine grained label.
- The authors are wrapping vanilla clustering technique with lots of fancy terms like semantic expansion. But once you read that section, it’s nothing else but clustering.
- Their Fine-Grained NER module is interesting to me. However, the most interesting part is not the technique but the hierarchical ontology. I understand their reason to manually label clusters instead of re-using some WordNet ontologies but I wonder if their clusters are as intuitive as WordNet.
- They should not sell their clustering as “knowledge base” because their clusters provide only “is-a” relation but a KB usually offers lots more!
- The rest of their modules are not interesting to me as most of them are outdated or underperforming the transformer models.
- 5: Transformative: This paper is likely to change our field. It should be considered for a best paper award.
- 4.5: Exciting: It changed my thinking on this topic. I would fight for it to be accepted.
- 4: Strong: I learned a lot from it. I would like to see it accepted.
- 3.5: Leaning positive: It can be accepted more or less in its current form. However, the work it describes is not particularly exciting and/or inspiring, so it will not be a big loss if people don’t see it in this conference.
- 3: Ambivalent: It has merits (e.g., it reports state-of-the-art results, the idea is nice), but there are key weaknesses (e.g., I didn’t learn much from it, evaluation is not convincing, it describes incremental work). I believe it can significantly benefit from another round of revision, but I won’t object to accepting it if my co-reviewers are willing to champion it.
- 2.5: Leaning negative: I am leaning towards rejection, but I can be persuaded if my co-reviewers think otherwise.
- 2: Mediocre: I would rather not see it in the conference.
- 1.5: Weak: I am pretty confident that it should be rejected.
- 1: Poor: I would fight to have it rejected.