1.This includes tools for scraping patent data from google patents.This is included in 'reponse_patents'.
2.A multi-dimensional classification algorithm using logistic regression and decision trees for ranking of a patent is provided in 'observation_analysis'.Data is split is 80:20 training test set randomly.Also the number of input variables is varied. Output of each of these combinations of the classifier on test data is saved.Test files compiles the results of each iteration in a single CSV file.
3. EPO data fetching engine