Help for categorizing RSS feeds!!

sasss52

Hi,
  I want to classify some RSS Feeds in four different category forma a specific location and after categorizing them I want to put them into different categorize folder according to the category. Can it be done by using RM? Or can someone give me some idea how to do this sort of categorization?

IngoRM

Hi,

assuming that you collect the RSS feeds and store them in a data format which can be read by RapidMiner: sure this can be done.

You can search for "text classification" here in this forum to get some information. You should also download the "Text" plugin and have a look at the samples there. And of course we also offer training courses for text classication / sentiment analysis / web mining etc.

Cheers,
Ingo

sasss52

HI Ingo,
  Thanks for ur reply. I already download the Text plugIn  and I run a text classification for some files and it is givings me the correct classification on the GIU. The thing I want to know is that, after the classification can i separate the file according to the their classification using RM or I have to write my own code to archive this or RM got a built in function to do this.
                                                                                                                          Cheers,

IngoRM

Hi,

yes, simply use the operators "ValueIterator" on the predicted label. This operator loops over all of the categories. Then use an "ExampleFilter" to keep only those of the examples where the predicted label is equal to the current value (stored in the macro "loop_value" which can be accessed by "%{loop_value}"). Then write down the subsets into different files / tables whatever, e.g. with the operator ExampleSetWriter using again the macro value of the current loop value in the filename.

Cheers,
Ingo

sasss52

Hi Ingo,
    Sorry for bothering you again and again. For classification I used "04_TextTraining.xml" example setup for my training and for test I use "05_TextModelApply.xml" setup. The thing I want to know now, where do I use  "ValueIterator" operator?is it after "Model Applier" using RM?? Is there any example that I can look for?? Can please give me some idea how to do this please!

TobiasMalbrecht

Hi,

if you want to separate the examples by their predicted class than you should place the [tt]ValueIterator[/tt] indeed behind the application of the prediction model, i.e. the [tt]ModelApplier[/tt]. Apart from that I think Ingo's explaination is quite comprehensive and should give you some idea. It might only be additionally noted that you should place the [tt]ExampleFilter[/tt] as a child of the [tt]ValueIterator[/tt]. The latter one is a meta operator applying its children for each of the values of the specified attribute. In each loop the macro [tt]loop_value[/tt] is set to the current attribute value. This macro can be accessed in the parameters of e.g. the [tt]ExampleFilter[/tt] operator as Ingo has described.

Hope that helps,
Tobias