The research on Latent Semantic Analysis (LSA) in the domain of natural language processing (NLP) shows the efficiency of this method (Landauer & Dumais, 1997) and nowadays the applicability in interactive semantic rich E- Learning contexts is interesting. Future applications may use LSA-techniques for automatic tutoring (Graesser et al., 1999) or automatic scoring of written essays in trainings, tests or MOOCs, which will be the way out of single-choice and multiple- choice tests in interactive learning environments. Text understanding is central in interactive learning environments. Therefore a written essay is a much better indicator than single or multiple choice questions and answers. The question of rating if a text has been understood is followed by the more important question of what have not been understood within the text, thus intelligent feedback can be given. LSA can regard, if knowledge has been decoded into a written essay or not. If text information is not included, it is still unclear whether the information was not understood or just not enough activated (Anderson, 1976).