A collection of various means for loading various different corpora used in NLP.
As per the standard Julia package installation:
Also, in the Pkg REPL, the package can be added with the add command, as:
pkg> add CorpusLoaders
For some corpus which we will say has type
it will have a constructior
path argument is a path to the files describing it.
That path will default to a predefined data dependency, if not provided.
The data dependency will be downloaded the first time you call
When the datadep resolves it will give full bibliograpghic details on the corpus etc.
For more on that like configuration details, see DataDeps.jl.
Each corpus has a function
This will return some iterator of data.
It is often lazy, e.g. using a
Channel, as many corpora are too large to fit in memory comfortably.
It will often be an iterator of iterators of iterators ...
Designed to be manipulated by using MultiResolutionIterators.jl.
The corpus type is an indexer for using named levels with MultiResolutionInterators.jl.
lvls(Corpus, :para) works.
Follow the links below for full docs on the usage of the corpora.