Code refactoring

[jdchoi77]

• Put elit/lemmatization and elit/token_tagger under elit/nlp.
• Create elit/model.py, move the classes in model/cnn_model.py and model/rnn_model.py to elit/model.py, and remove elit/model.
• Remove elit/state.py (we no longer use it?)
• Remove resources/morph_analyzer and rename resources/lemmatizer to resources/morph_analyzer. You probably need to reconfigure the resource path for EnglishMorphAnalyzer (we probably don't need EnglishLemmatizer any more).