1- Run the project
1-Download a doc2vec model (https://github.com/jhlau/doc2vec) and unpack it in the folder 2-run the file script.py 3-follow the instructions to select the target, the features and the cardinality of the training set 4-The predictions will be saved in the partial results folder
Run fast_test_script.py if you prefer to skip the interactive part, but you have to modify the code in order to select the features.
NOTE: you need to manually comment the PCA and feature selection functions in dataExtractor.py if you don't want to use them.
The dataset needs to be in the .csv format. It is composed as follows: 1- User Stories (String) 2- Business Value (String) 3- User Story Elaboration (String) 4- Definition of done (String) 5- Expected output (String) 6- LOC (int) 7- New classes (int) 8- Changed classes (int) 9- Effort (int) 10- N. Unit Test (int) 11- Entropy (int) 12- Services (Matrix of int)