Size: 100 pairs
Task: for a given pair of words one should determine what word is closer to the maximum value (e.g. fast is closer to max and slow is closer to min).
Anntotation: each pair is annotated in the following way:
fast 1, slow 0
where 1 and 0 are for maximum and minimum respectively.
Conatains only andjectives which are used for quantity.
Size: 38 pairs
File contains 91 question beginnings for English, sorted in alphabetic order.
later
There are some data sets for the task of determining whether the given adjective is an attribute for a noun or not (e.g. fast is an attribute for speed but not for weight).
Created via WordNet in NLTK. Contains about 700 groups.
Organized in the following way:
Noun : its attributes
Example:
sureness:incertain, certain, sure, uncertain, unsure
Inverted set from the previous one. Countains about 860 groups.
Organized in the following way:
Adjective: nouns which it belongs to
Example:
long:duration, length
These sets can be used for training and testing your models.
Both sets contain 200 samples, 3-5 candidates in each sample.
Organized in the following way:
Adjective:Right noun:set of noun candidates
Example:
deep:depth:orient,fervor,depth,likelihood