running RM process as deamon
I'm writing a text classification process. A SVM model and wordlist are pre-generated. Since loading the model and wordlist takes some time, i'm wondering what's the best way to deamonize the process. That is, loading model/wordlist just once, process any incoming documents as needed.