corenlp_annotate: Annotate a string.

In PolMine/bignlp: Fast and Memory-Efficient Annotation of Big Corpora

Description

Use CoreNLP to annotate strings.

Usage

corenlp_annotate(x, ...)

## S4 method for signature 'data.table'
corenlp_annotate(
  x,
  corenlp_dir = getOption("bignlp.corenlp_dir"),
  properties,
  purge = TRUE,
  threads = 1L,
  progress = TRUE,
  verbose = TRUE
)

## S4 method for signature 'character'
corenlp_annotate(
  x,
  corenlp_dir = getOption("bignlp.corenlp_dir"),
  properties,
  byline = NULL,
  output_format = "json",
  threads = 1L,
  progress = TRUE,
  preclean = TRUE,
  verbose = TRUE
)

Arguments

x	Either a data.table (required to have the columns 'doc_id' and 'text'), or a character vector with input file(s), or a directory. If input is a directory, all files in the directory are processed. Files are assumed to be tsv files with two columns ('doc_id' and 'text').
...	Further arguments.
corenlp_dir	The directory where corenlp resides.
properties	A properties file to configure annotator.
purge	A logical value, whether to preprocess input.
threads	An integer value.
progress	Logical, whether to show progress bar.
verbose	Logical, whether to output messages.
byline	Logical, whether to process files in a line-by-line manner.
output_format	The output generated, either "json" (default), "txt", or "xml".
preclean	Logical, whether to preprocess string.

Details

If argument threads is 1, the tagging result is returned, if output is NULL. If threads is higher than 1, output should be a directory where tagging results will be stored as NDJSON files.

Value

The target files will be returned, so that they can serve as input to corenlp_parse_ndjson.

Examples

library(data.table)
reuters_txt <- readLines(system.file(package = "bignlp", "extdata", "txt", "reuters.txt"))
reuters_dt <- data.table(doc_id = 1L:length(reuters_txt), text = reuters_txt)

props <- corenlp_get_properties_file(lang = "en", fast = "TRUE")
y <- corenlp_annotate(
  x = reuters_dt,
  properties = props,
  corenlp_dir = corenlp_get_jar_dir(),
  progress = FALSE
)

PolMine/bignlp documentation built on Jan. 29, 2021, 1:14 a.m.