AIAS1/2_nlp_sdks/lexical_analysis/lac_sdk

Download the model and place it in the models directory

• Link: https://github.com/mymagicpower/AIAS/releases/download/apps/lac.zip

Text-Lexical Analysis SDK [Chinese]

The lexical analysis model can complete Chinese word segmentation, Part-of-Speech tagging, and Named Entity Recognition tasks as a whole. Part-of-Speech tagging: -n Common Noun -f Locative Noun -s Place Noun -t Time -nr Person Name -ns Place Name -nt Organization Name -nw Work Name -nz Other Proper Nouns -v Verb -vd Verb Adverb -vn Verb Noun -a Adjective -ad Adverbial Adjective -an Nominal Adjective -d Adverb -m Quantifier -q Quantity Word -r Pronoun -p Preposition -c Conjunction -u Auxiliary -xc Other Particles -w Punctuation -PER Person Name -LOC Place Name -ORG Organization Name -TIME Time

SDK Algorithm:

Model network structure:

• Paper：
  https://arxiv.org/abs/1807.01882

Run Example-LacExample

After a successful run, you should see the following information on the command line:

...
[INFO ] - input Sentence: 今天是个好日子
[INFO ] - Words : [今天, 是, 个, 好日子]
[INFO ] - Tags : [TIME, v, q, n]

Open source algorithm

1. Open source algorithm used by SDK

• PaddleNLP

2. How to export the model?

• how_to_create_paddlepaddle_model