This is a rather broad question, so I apologize in advance.
I have been trying to find
uniquely named datasets that are widely used in different domains. Eg, in the following sentence:
We used two corpora for our analysis: hospital discharge summaries from 1991 to 1997 from the Columbia-Presbyterian Medical Center, and the January 1996 part of the Wall Street Journal corpus from the Penn TreeBank [Marcus et al].The phrase the Wall Street Journal corpus is a valid example for my purpose, but the phrase hospital discharge summaries from 1991 to 1997 from the Columbia-Presbyterian Medical Center is not, while both of them refer to datasets.
Some more examples are:
- Brown corpus (NLP)
- Enron email dataset (NLP, data mining)
- MNIST handwritten digit database (computer vision, machine learning)
Note that many papers can refer to the last dataset as just "MNIST", instead of writing the full name. Hence, the "uniquely named" criterion is not very strict. But as a researcher in the domain, you should be able to put it into an equivalence class, i.e., you should understand that MNIST and Brown corpus are two distinct entities (similar to the named entity disambiguation problem).
In NLP, data mining and computer vision, there is a specific pattern to this, i.e., people usually refer to them as dataset or corpus. One of my friends in astrophysics told me that they are referred as catalogue in their domain.
Given this information, my question is:
Is there a pattern in different domains to refer to datasets? I am interested mostly in biology and economics.
More importantly, is there a curated list of such named datasets?
I looked at such a curated list, but it has a problem. Most links refer to the data repositories as opposed to specific datasets. Eg, the link associated with the phrase Blogger corpus links to Blog Authorship Corpus, which is an actual dataset. But,
CERN Open Data Portal links to a repository of datasets.
Sorry for the long post, but any feedback would be appreciated.