During Week 6, I met my external advisor to discuss the progress of my project. He gave me a few tips for making my code more readable and more organized. In the beginning of the week, I came up with the idea to save all the features in files to save time running my FeatureExtractor class with the help of my advisor and dad. I saved all the features into files using a FileOutputStrean and a ObjectOutputStream and used a FileInputStrean and a ObjectInputStream to read all the features from its file.
After saving all the article features into files, I started to train my classifier with 200 articles and test with 20 articles and got a result of 6/20 of the testing articles correct. I then tried to increase the number of training articles to 500 and got a result of 10/20 of the testing articles correct. Instead of running for 24 hours, the classifier now runs for a few minutes, which allows me to do more testing on the accuracy of my classifier. For next week, I will be testing my classifier even more and finding more articles to use as data for my classifier.
Thanks for reading!