The Altair Community is migrating to a new platform to provide a better experience for you. In preparation for the migration, the Altair Community is on read-only mode from October 28 - November 6, 2024. Technical support via cases will continue to work as is. For any urgent requests from Students/Faculty members, please submit the form linked here
Starting fresh on Mac. HELP!!
kebuenokebueno
Member Posts: 1 
Learner III
Basically I tried to process a file that was way too big to process it (about 40K text data instances). I tried cancelling the upload, but I must have done something wrong because now, everytime I try starting up the program by running the rapidminer.jar, it just leaves me out after the second click.
Please note that when I try starting the program, on the very first menu I have an empty bar on my right, as if I was loading something, but is empty. When I try clicking something on the main menu, it doesnt matter what, it leaves me out.
Now, does anybody know which file/folder/line in my computer has to do with that part of the configuration of the program? It is very frustrating
I have snow leopard 10.6 and 4G of RAM
That was another question: I want to do text classification with about 40-60K training set instances. Every time there is a new text to classify, there are all those instances to probabilistically support the classification. So far there are about 90 different "text-types" these texts could belong to. Looking at theories and accuracies, I would like to do multiclass SVM.
Am I being a bit delusional? Should I go for something simpler? I am opened to all suggestions.
Much appreciated!!!
Please note that when I try starting the program, on the very first menu I have an empty bar on my right, as if I was loading something, but is empty. When I try clicking something on the main menu, it doesnt matter what, it leaves me out.
Now, does anybody know which file/folder/line in my computer has to do with that part of the configuration of the program? It is very frustrating
I have snow leopard 10.6 and 4G of RAM
That was another question: I want to do text classification with about 40-60K training set instances. Every time there is a new text to classify, there are all those instances to probabilistically support the classification. So far there are about 90 different "text-types" these texts could belong to. Looking at theories and accuracies, I would like to do multiclass SVM.
Am I being a bit delusional? Should I go for something simpler? I am opened to all suggestions.
Much appreciated!!!
0
Answers
usually RapidMiner creates a .RapidMiner5 directory in your user directory. If you delete it, everything should be back on defaults.
It is possible to train these amount of texts, but I think you will have to go for a faster computer with more memory...
And don't expect a to high classification rate. Always keep in mind that the default classification rate is around 2% (if classes are equally frequent), so even 40% correct classification would be an improvement of factor 40...
If you need advice or practical help for classifying these texts, please don't hesitate to contact us. We have some experience with these amounts of data.
Greetings,
Sebastian