Advantages of the Explorer Interface

The strength of this interface lies in its simplicity, but it does this without hiding a lot of the functionality that is possible in Weka. All of the various learning algorithms filters and visualization tools are available and accessible in this interface.

Disadvantages of the Explorer Interface

The main weakness of the explorer is that is designed to work with one learning model at a time and there is no way to compare different models as to their performance. This problem, however, is addressed in the Experimenter and KnowledgeFlow interfaces.

Another more serious problem is the lack of a data editor. In our example, we were forced to use filters to convert attributes that were nominal but were detected as numeric. Explorer already detects the distinct values in these attributes and what is needed is a way of converting these attributes without having to use a filter to do it. It would also be nice to be able to change the labels on nominal attributes. Being able to change the name of a model that has be trained, or a dataset that has just been processed by a filter would also make the output generated easier to read.