shubhradeepnandi / Bible_Dataset

Bible_text_classification

TEXT CLASSIFICATION ON BIBLE DATA SET INTO OLD AND NEW TESTAMENT:

--> There are basically four folders:

1. The first folder is bible_data_set: Contains the input dataset

2. Machine Learning is the folder that contains two python notebooks and their respective models in the internal model folder.

   a) training_word2vec.ipynb : Trained on all samples using word2vec Vectorization and SVM classification algorithm.

     contains the answer to first two questions :
   
     1) Classification into old and new testament : Accuracy --> 78%
     2) The cosine distance between light and dark comes out to be 0.5777796
   

   b) training_tfidf.ipynb : Trained on all samples using TFIDF Vectorization and SVM classification algorithm.

     contains the answer to first question :
   
     1) Classification into old and new testament : Accuracy --> 92.03729503804523%
   

3. Deep Learning is the folder that contains 1 python notebook and its respective model.

   a) NLP-BIBLE.ipynb : Trained on all samples using word2vec Vectorization and LSTM classification.

     contains the answer to first two questions :
   
     1) Classification into old and new testament : Accuracy --> 74.42%
     2) The cosine distance between light and dark comes out to be 0.0026299357414245605
   

4. Nouns_seggregation is the fourth folder that contains 1 python notebook.

   a) noun_seggregation.ipynb : Creates a data_with_noun.csv file in the bible_data_set folder with first 1000 sentences and their respective nouns in the noun column. Due to processing time problem ,only 1000 sentences have been processed.