APERTIUM-TAGGER(1)

✨ APERTIUM-TAGGER(1) ✨

Okay bestie, this is the manual page for apertium-tagger 💅 It's the part-of-speech tagger and trainer for Apertium, so it can tag your text or train a model, depending on which flags you throw at it 🧚

---

🌸 NAME

apertium-tagger is the part-of-speech tagger and trainer for Apertium 💖

---

🦋 SYNOPSIS

apertium-tagger [options] -g serialized_tagger [input [output]]
apertium-tagger [options] -r iterations corpus serialized_tagger
apertium-tagger [options] -s iterations dictionary corpus tagger_spec
                serialized_tagger tagged_corpus untagged_corpus
apertium-tagger [options] -s 0 dictionary tagger_spec serialized_tagger
                tagged_corpus untagged_corpus
apertium-tagger [options] -s 0 -u model serialized_tagger tagged_corpus
apertium-tagger [options] -t iterations dictionary corpus tagger_spec
                serialized_tagger

---

💜 DESCRIPTION

Okay so apertium-tagger is the app that handles the Apertium part-of-speech tagger, and whether it's training or tagging depends entirely on the calling options 🔥 Lowkey important detail: it only reads from standard input if you use --tagger or -g 🙈

---

🎀 MODES

• -g, --tagger — Tags your input text using the Viterbi algorithm 🫶
• -r n, --retrain n — Retrains the model with n extra Baum-Welch iterations (unsupervised). Not compatible with -u (--unigram)! 💀
• -s n, --supervised n — Starts the parameters off against a hand-tagged text (supervised) using the maximum likelihood estimate method, then runs n iterations of the Baum-Welch training algorithm (unsupervised) ✨ The CRP argument can only be left out when n = 0.
• -t n, --train n — Initializes the parameters with Kupiec's method (unsupervised), then runs n iterations of the Baum-Welch training algorithm (unsupervised) 🌈

---

🧚 MODELS

• -u, --unigram=MODEL — Uses the unigram algorithm MODEL from https://coltekin.net/cagri/papers/trmorph-tools.pdf 📄
• -w, --sliding-window — Uses the Light Sliding Window algorithm 🪟
• -x, --perceptron — Uses the averaged perceptron algorithm 🍓

---

⭐ OPTIONS

• -d, --debug — Prints errors (if there are any) or debug messages while it's running 🔍
• -e, --skip-on-error — Used with -xs to skip certain types of errors in the training corpus 🙃
• -f, --first — Used with -g (--tagger) so the tagger gives you all the lexical forms of each word, with the chosen one sitting in the first place (after the lemma) 👑
• -m, --mark — Marks the disambiguated words 💅
• -p, --show-superficial — Prints the superficial form of the word alongside the lexical form in the output stream 🦋
• -z, --null-flush — Used with -g (--tagger) to flush the output after getting each null character 🌊
• --help — Displays a help message 🫶

---

📁 FILES

These are the kinds of files each option uses 🗂️

• dictionary — Full expanded dictionary file 📖
• corpus — Training text corpus file 📝
• tagger_spec — Tagger specification file, in XML format 🧾
• serialized_tagger — Tagger data file, built during training and used while tagging 💾
• tagged_corpus — Hand-tagged text corpus 🏷️
• untagged_corpus — Untagged text corpus, the morphological analysis of the hand-tagged corpus, used together with the -s option 🧩
• input — Input file, standard input by default 📥
• output — Output file, standard output by default 📤

---

🔗 SEE ALSO

Go check these out too 👀✨

• apertium(1), lt-comp(1), lt-expand(1), lt-proc(1)

---

📜 COPYRIGHT

Copyright © 2005, 2006 Universitat d'Alacant / Universidad de Alicante. This is free software 🕊️ You're allowed to redistribute copies of it under the terms of the GNU General Public License: https://www.gnu.org/licenses/gpl.html.

---

🐛 BUGS

Many... lurking in the dark and waiting for you! 💀👻

---

Okay that's the whole tea ☕💖 Source: Apertium, February 22, 2021, APERTIUM-TAGGER(1). Stay iconic, bestie! 🌈✨