LoginSign Up
← Back to all sparks
N

nametagger

AI-ASSISTANTS
Velocity0.0

Named Entity Recognition in Texts using NameTag

github.com/bnosac/nametagger ↗

An R binding to NameTag that has not gained a feature since its 2020 debut.

rnlpnamed-entity-recognitionbindingsmaintenance
◆Current state
nametagger wraps UFAL's NameTag for named entity recognition in R, letting users apply and train NER models on tokenized text. Every release after the initial 0.1.0 is compiler or CRAN conformance work: misaligned-address and UBSan reports, a C++20 declaration fix for persistent_unordered_map, dropping C++11, and a sprintf swap. The R-level API has not moved.
◆Where it's heading
The package is maintained as part of a family of bnosac NLP bindings that are updated together — the same C++20 persistent_unordered_map fix appears in udpipe within days, and the C++11 drops across the family landed in the same sweep. Releases are triggered by CRAN's checks, not by NameTag's own development.
◆Prediction
Expect the next release to be whichever compiler conformance issue CRAN raises next, most likely arriving alongside matching fixes in the sibling packages.

◆Recent moves

  1. 6mo ago

    Misaligned address and UBSan fixes

    Fixes a misaligned address load and UBSan messages flagged by CRAN. The same class of fix udpipe received in its own 0.8.13 release.

    View source ↗
  2. 6mo ago

    persistent_unordered_map declaration fixed for C++20

    Fixes the persistent_unordered_map declaration for C++20, plus authorship metadata and an http to https documentation URL. udpipe shipped the identical C++20 fix days later, confirming these packages are maintained as one family.

    View source ↗
  3. 2y ago

    C++11 dropped; std::iterator removed from UTF headers

    Silences a false-positive valgrind report, drops the C++11 specification, and pulls upstream nametag fixes removing std::iterator from the UTF handling headers. Vendored-code upkeep.

    View source ↗
  4. 3y ago

    snprintf swap for M1 Mac check note

    Replaces sprintf with snprintf for an R CMD check note on M1 Macs and adds a README example. The same M1 fix appears across the sibling packages.

    View source ↗
  5. 5y ago

    udpipe moved from Imports to Suggests

    Moves udpipe from Imports to Suggests and drops crfsuite, making the tokenization dependency optional and the example conditional. A packaging change rather than a functional one.

    View source ↗
  6. 6y ago

    Initial release wrapping UFAL NameTag

    The initial release, wrapping a pinned commit of UFAL's NameTag. This is the only entry in the record that added capability; everything since keeps it compiling.

    View source ↗