Some observations: - **Currency:** A currency symbol should be related to a numeric amount via [`nummod`](http://universaldependencies.org/u/dep/nummod.html) (e.g. in "$ 40"), but in the English data [it is sometimes `compound`](http://bionlp-www.utu.fi/dep_search/query?search=%22%24%22%20%21%3Enummod%20_&db=UD_English-v2&case_sensitive=True&hits_per_page=50). - **Dates:** [`flat`](http://universaldependencies.org/u/dep/flat.html) includes examples of dates, but the English data has `nummod` (e.g., [“March 28”](http://bionlp-www.utu.fi/dep_search/query?search=March&db=UD_English-v2&case_sensitive=True&hits_per_page=10)) - **Times:** There is no apparent guideline for "10:00 pm" or "10 o'clock". In the English data, "a.m." and "p.m." are attached as `nummod`. - **Addresses:** It is not clear what the standard should be—`flat`?—for expressions like "100 Main St. NW Apt. 10". - **Spelled-out numbers:** [`flat`](http://universaldependencies.org/u/dep/flat.html) gives the example "four thousand", but the English data has `compound` (e.g., [these examples with "million"](http://bionlp-www.utu.fi/dep_search/query?search=_%20%3Ccompound%20million&db=UD_English-v2&case_sensitive=True&hits_per_page=10))
Some observations:
Currency: A currency symbol should be related to a numeric amount via
nummod(e.g. in "$ 40"), but in the English data it is sometimescompound.Dates:
flatincludes examples of dates, but the English data hasnummod(e.g., “March 28”)Times: There is no apparent guideline for "10:00 pm" or "10 o'clock". In the English data, "a.m." and "p.m." are attached as
nummod.Addresses: It is not clear what the standard should be—
flat?—for expressions like "100 Main St. NW Apt. 10".Spelled-out numbers:
flatgives the example "four thousand", but the English data hascompound(e.g., these examples with "million")