Java Text Classification using Naive Bayes

I require a Java command line program that can perform text classification (binary) using Naive Bayes (maybe include TF-IDF if possible). This is very similar to a Spam Filter (i.e. Ham or Spam). Here are several examples [login to view URL] [login to view URL] The program should use Java 8 and use a open source machine library like Apache MLLib, Apache Mahout, DeepLearning4J, [login to view URL], JATECS, or some other (do not use WEKA). If you know of another library that you want to use, let me know. The program should have the following functionality: - Reads in a CSV file (the first column will be the binary outcome, and the second column will be the Text) - Process Data File (like stemming, remove stop words, etc.) - Train on 70% of data; Test on 30% of the data - Output Result should be a Confusion Matrix and an Accuracy Calculation. If a sentence is provided then should print out the predicted binary outcome. The java program should take the following parameters: Java -jar [login to view URL] [login to view URL] “sentence” Note: “sentence” parameter is optional and should print the predicted binary outcome. I have provided 3 sample data files here: [login to view URL] I would prefer someone that has done text classification before. If you can suggest other techniques let me know. Deliverables include the following: - Source code with documentation - Jar file