NLPServer implements a server listening http requests on different routes, to query different modules of the KotlinNLP library.
NLPServer is part of KotlinNLP.
Run the server simply passing configuration parameters by command line to the RunServerKt script.
This is the help command output:
usage: [-h] [-p PORT] [-t TOKENIZER_MODELS_DIRECTORY]
[-l LANGUAGE_DETECTOR_MODEL] [-c CJK_TOKENIZER_MODEL]
[-f FREQUENCY_DICTIONARY] [-m MORPHOLOGY_DICTIONARY]
[-e PRE_TRAINED_WORD_EMB] [-n NEURAL_PARSER] [-x FRAME_EXTRACTOR]
[-s HAN_CLASSIFIER] [-d LOCATIONS_DICTIONARY]
optional arguments:
-h, --help show this help message and exit
-p PORT, --port PORT the port listened from the server
-t TOKENIZER_MODELS_DIRECTORY, the directory containing the
--tokenizer-models-directory TOKENIZER_MODELS_DIRECTORY serialized models of the neural
tokenizers (one per language)
-l LANGUAGE_DETECTOR_MODEL, the filename of the language
--language-detector-model LANGUAGE_DETECTOR_MODEL detector serialized model
-c CJK_TOKENIZER_MODEL, the filename of the CJK neural
--cjk-tokenizer-model CJK_TOKENIZER_MODEL tokenizer model used by the
language detector
-f FREQUENCY_DICTIONARY, the filename of the frequency
--frequency-dictionary FREQUENCY_DICTIONARY dictionary used by the language
detector
-m MORPHOLOGY_DICTIONARY, the directory containing the
--morphology-dictionary MORPHOLOGY_DICTIONARY morphology dictionaries used by
the parser (one per language)
-e PRE_TRAINED_WORD_EMB, the directory containing the
--pre-trained-word-emb PRE_TRAINED_WORD_EMB pre-trained word embeddings
files, one per language (the file
name must end with the ISO 639-1
language code)
-n NEURAL_PARSER, --neural-parser NEURAL_PARSER the directory containing the
serialized models of the neural
parsers (one per language)
-x FRAME_EXTRACTOR, --frame-extractor FRAME_EXTRACTOR the directory containing the
serialized models of the frame
extractors, one per domain
-s HAN_CLASSIFIER, --han-classifier HAN_CLASSIFIER the directory containing the
serialized models of the HAN
classifier, one per domain
-d LOCATIONS_DICTIONARY, the filename of the serialized
--locations-dictionary LOCATIONS_DICTIONARY locations dictionary
<dependency>
<groupId>com.kotlinnlp</groupId>
<artifactId>nlpserver</artifactId>
<version>0.6.3</version>
</dependency>The KotlinNLP RESTful APIs are designed with the OpenAPI Specification.
Install the API client running install_client.sh.
Try some examples of usage of the KotlinNLP APIs running the files in the examples folder.
This software is released under the terms of the Mozilla Public License, v. 2.0
We greatly appreciate any bug reports and contributions, which can be made by filing an issue or making a pull request through the github page.