def get_lda_model_hyperp():
return {
mh.LDA_Model_Hyperp.LDA_MODEL_N_COMPONENTS.value: 20,
mh.LDA_Model_Hyperp.LDA_MODEL_RANDOM_STATE.value : 2,
mh.LDA_Model_Hyperp.VECTORIZER_NGRAM_RANGE.value: (1,1),
mh.LDA_Model_Hyperp.VECTORIZER_MAX_FEATURES.value: 200,
mh.LDA_Model_Hyperp.VECTORIZER.value : TfidfVectorizer(stop_words='english', use_idf=True, smooth_idf=True),
mh.LDA_Model_Hyperp.VECTORIZER_TOKENIZER.value : tok.PorterStemmerBased_Tokenizer()
}
def get_cust_w2v_hyperp():
return {
mh.WordVec_Model_Hyperp.TOKENIZER.value:
tok.PorterStemmerBased_Tokenizer(),
mh.WordVec_Model_Hyperp.WORD_EMBEDDING.value:
'CUSTOMIZED',
mh.WordVec_Model_Hyperp.GEN_NAME.value:
'cust_wordvector'
}
def get_bm25_model_hyperp():
return {
mh.BM25_Model_Hyperp.TOKENIZER.value : tok.PorterStemmerBased_Tokenizer()
}