nltkopennlpdeeplearning4jocldkpro-core

Natural language processing tools for generating OCL


please I have a university project called (automatic generation of OCL

constraints) and my supervisor asked me to choose a tool from this list of tools

for natural language processing:

Apache OpenNLP, Deeplearning4j, ChatScript, DELPH-IN, DKPro Core, general architecture text engineering GATE, Gensim, LinguaStream, Mallet (software project), Modular Audio Recognition Framework, MontyLingua, Natural Language Toolkit, SpaCy, UIMA.

what would be the easiest to implement and which one would be most suitable for my future work?

else any propositions!


Solution

  • You should beware of a fundamental contradiction.

    Users of OCL typically expect 100% accuracy.

    Users of NLP are generally very pleased to achieve greater than 95% accuracy.

    You must therefore restrict your domain to where 95% accuracy is acceptable, and this may influence your language choices.