Helper function to extract and categorize tags from a loaded Flair SequenceTagger model. The tags are grouped into categories such as person, organization, location, and miscellaneous.
Value
A list of tags grouped by category:
- all
Complete list of all available tags
- special
Special tags like
, O, , - person
Person-related tags (e.g., B-PER, I-PER)
- organization
Organization tags (e.g., B-ORG, E-ORG)
- location
Location tags (e.g., B-LOC, S-LOC)
- misc
Miscellaneous entity tags
Details
The tags follow the BIOES (Begin, Inside, Outside, End, Single) scheme:
B-: Beginning of multi-token entity (e.g., B-PER in "John Smith")
I-: Inside of multi-token entity (e.g., I-PER in "John Smith")
O: Outside of any entity
E-: End of multi-token entity
S-: Single token entity (e.g., S-LOC in "Paris")
See also
load_tagger_ner for loading the NER model
Examples
if (FALSE) { # \dontrun{
# Load a NER model
tagger <- load_tagger_ner("flair/ner-english-large")
# Extract all tags
tags <- get_tagger_tags(tagger)
# Access specific tag categories
print(tags$person) # All person-related tags
print(tags$location) # All location-related tags
# Example usage with text annotation
# B-PER I-PER O S-ORG
# "John Smith works at Google"
# B-LOC E-LOC O B-ORG E-ORG
# "New York is United Nations headquarters"
# Use tags to filter entities
person_entities <- results[tag %in% tags$person]
org_entities <- results[tag %in% tags$organization]
} # }