This repository hosts the codes necessary to reproduce the results of our approach towards Morphological analysis of Hindi words.
-
data_preprocessing contains the files necessary to parse and load the data.
- The file load_data.py contains functions callabe at the time of training and testing the models.
-
models hosts the various models implemented for predicting the rootword as well as the tag features.
-
result_visualization contains the files necessary for plotting the performance metric curves.
-
curve_outputs contains the graphical outputs of each model in terms of average precision-recall scores for each classes.
-
data_and_train_info hosts the stats about the Hindi TreeBank dataset that we extracted as per our needs along with the information regarding training of the models and their performance.