How ye-kyaw-thu/mypos works
myPOS is a manually segmented and POS-tagged Myanmar corpus, accompanied by validation, tag conversion, and research model implementations. The user-facing research workflow is to prepare corpus examples, train or decode a tagger, and evaluate its predictions. The sampled NCRF++ code shows data preparation and model composition; the README documents corpus segmentation and the tag set. The graph preserves the listed model families and conversion utility, but their detailed runtime links are not inferred where source excerpts are unavailable.
Main entry points: 1_NCRFPP_with_otest_50%_ASEAN/main.py NCRF++ 50% run · 2_NCRFPP_with_otest_100%_ASEAN/main.py NCRF++ 100% run · data.py Data preparation · seqlabel.py Sequence labeler
Components 29
Corpus and tags
- myPOS corpus
Tagging experiments
- Draft HMM evaluation
corpus-draft-ver-1.0/model/3gHMM/evaluate.py - Draft CRF evaluation
corpus-draft-ver-1.0/model/crf/evaluate.py - KyTea evaluation
corpus-draft-ver-1.0/model/kytea/evaluate.py - Low-resource evaluation
corpus-draft-ver-1.0/model/low-resource-pos-tagging-2014/evaluate.py - MaxEnt evaluation
corpus-draft-ver-1.0/model/maxent/evaluate.py - Draft RDR evaluation
corpus-draft-ver-1.0/model/rdr/evaluate.py - RDR evaluations
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/1_RDR/1_RDR_with_otest_50%_ASEAN_Corpus/evaluate.py - CRF evaluations
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/2_CRF/1_CRF_with_otest_50%_ASEAN_Corpus/evaluate.py - HMM evaluations
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/3_HMM/1_HMM_with_otest_50%_ASEAN_Corpus/evaluate.py - NCRF++ 50% run
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/main.py - NCRF++ 100% run
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/2_NCRFPP_with_otest_100%_ASEAN/main.py - Version 3 HMM evaluation
corpus-ver-3.0/models/3gHMM_mypos-ver3/evaluate.py - Version 3 NCRF evaluation
corpus-ver-3.0/models/NCRFpp_mypos-ver3/evaluate.py - Trained NCRF evaluation
corpus-ver-3.0/models/NCRFpp_mypos-ver3/trained-models/wordCNN-CRF-charLSTM/evaluate.py - Version 3 CRF evaluation
corpus-ver-3.0/models/crf_mypos-ver3/evaluate.py - Version 3 RDR evaluation
corpus-ver-3.0/models/rdr_mypos-ver3/evaluate.py
NCRF++ internals
- Data preparation
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/utils/data.py - Instance encoding
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/utils/functions.py - Token alphabets
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/utils/alphabet.py - Sequence labeler
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/seqlabel.py - Sentence classifier
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/sentclassifier.py - Word sequence encoder
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/wordsequence.py - Word representations
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/wordrep.py - CRF decoder
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/crf.py - Tag metrics
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/utils/metric.py
Corpus utilities
- Corpus validator
corpus-ver-3.0/etc/pos_corpus_validator.py - UPOS converter
myPOS-to-universalPOS/mypos2upos.py
Other components
- NLP researcher
Connections 18
- NLP researcher → myPOS corpus: uses
- NLP researcher → NCRF++ 50% run: runs experiments
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/main.py - NCRF++ 50% run → Data preparation: prepares data
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/main.py - NCRF++ 50% run → Sequence labeler: uses labeler
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/main.py - NCRF++ 50% run → Sentence classifier: uses classifier
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/main.py - NCRF++ 50% run → Tag metrics: scores predictions
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/main.py - NCRF++ 100% run → Data preparation: prepares data
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/2_NCRFPP_with_otest_100%_ASEAN/main.py - NCRF++ 100% run → Sequence labeler: uses labeler
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/2_NCRFPP_with_otest_100%_ASEAN/main.py - NCRF++ 100% run → Sentence classifier: uses classifier
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/2_NCRFPP_with_otest_100%_ASEAN/main.py - NCRF++ 100% run → Tag metrics: scores predictions
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/2_NCRFPP_with_otest_100%_ASEAN/main.py - Data preparation → Token alphabets: uses alphabets
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/utils/data.py - Data preparation → Instance encoding: encodes instances
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/utils/data.py - Sequence labeler → Word sequence encoder: encodes sequence
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/seqlabel.py - Sequence labeler → CRF decoder: decodes tags
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/seqlabel.py - Sentence classifier → Word sequence encoder: encodes sentence
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/sentclassifier.py - Word sequence encoder → Word representations: builds representations
corpus-ver-3.0/iSAI-NLP2020-paper-experiment/4_NCRFPP/1_NCRFPP_with_otest_50%_ASEAN/model/wordsequence.py - myPOS corpus → Corpus validator: validates corpus
- myPOS corpus → UPOS converter: converts tags