Takes a text input file and by default, produces a tab-delimited csv output file. Output columns do not have a header row, but are always arranged the same way in three columns.

ner.ps1

<#
    Requires Stanford NER, Java 1.8+
    formats = slashTags, inlineXML, xml, tsv, tabbedEntities 
#>


param(
    [Parameter(Mandatory=$true,Position=0)]
    [string]$file,
    [Parameter(Mandatory=$true,Position=1)]
    [string]$outfile,
    [string]$format = "tabbedEntities",
    [string]$model = "3class",
    [string]$jardir = "..\stanford-ner"
)

$classifier = switch ($model) {
    3class {"english.all.3class.distsim.crf.ser.gz"; break}
    4class {"english.conll.4class.distsim.crf.ser.gz"; break}
    7class {"english.muc.7class.distsim.crf.ser.gz"; break}
    default {"english.all.3class.distsim.crf.ser.gz"; break}
}

java -mx1500m -cp (join-path $jardir stanford-ner.jar) edu.stanford.nlp.ie.crf.CRFClassifier -loadClassifier (join-path $jardir classifiers\$classifier) -outputFormat $format -textFile $file > $outfile