In this section, we will discuss some common preprocessing and transformation steps that are done in most text mining processes. The general concept is to convert the documents into structured datasets with features or attributes that most Machine Learning algorithms can use to perform different kinds of learning.
We will briefly describe some of the most used techniques in the next section. Different applications of text mining might use different pieces or variations of the components shown in the following figure:
Figure 10: Text Processing components and the flow
One of the first steps in most text mining applications is the collection of data in the form of a body of documents—often referred to as a corpus in the text mining world. These documents can have predefined categorization associated with them or it can simply be an unlabeled corpus. The documents can be of heterogeneous formats or standardized...