Mining Complex Data

Traditional data mining focuses on examples given in a relatively simple attribute value format. That means, that every example has one specific value for each of the possible attributes. One of our research interests is the focus on data which is not easily transformable into such a simple format. Examples could have a variable number of attributes or more than just one single value for an attribute. Furthermore, some data can be best described by sequential string of even graphs or are arriving at massive amounts, such that they are impossible to be stored and analyzed as fast as the data arrives. In these sections, we give a brief overview of such data and our research focus in this fields.