Oracle Text mining in general does support Traditional and Simplified Chinese.
The Oracle Data Miner supports Text Mining in Build Text node.
In Build Text node, you can specify the language setting as Traditional or Simplified Chinese. You may also add Chinese stop words. The Build Text node is used to tokenize the input Text data (data preparation), which can then be processed by the Model build.
In the next JDev release (3.3), the text mining operation is imbedded in the Build process, so it's more tightly integrated.