Important files:
- cnn/train.py : To train CNN
- cnn/eval.py : Evaluate model
- nb_classifier.py : Naive Bayes classifier
- data_preprocessor.py : Pre process data
- data_tfidf.py : Calculate TF-IDF scores and find important words.
- nbt_body_extracter.py : Clean and Extract plain text content from Navbharattimes News Archieve.
- crawler.py : To download News Articles and extract required data like title, body, category, etc. (Customized for Navbharattimes)
- Summarization/tensorflow-glove/hindi-glove.py : Find glove embeddings
- Summarization/scr/train.py : Train RNN
- Summarization/src/predict.py : Evaluate model on BLEU score
Python version: Python 3
Python packages:
- Tensor flow 1.0
- sklearn (Scipy 0.18)
- numpy 1.12
- BeautifulSoup 4
- requests 2.13
- scipy
- keras 1.2.2
- Theano
- Levenshtein
- h5py
- nltk
- pandas
- seaborn
- matplotlib
- tqdm