KANSAS is an interdisciplinary project that touches upon the fields of adult education, computer linguistics, research on teaching, and cognitive psychology. The goal of the project is to develop, test, and evaluate a user-oriented, web-based search engine (KANSAS), which finds authentic written text samples for language based on their individual learning skills and requirements.

The search engine enables teachers of literacy courses to search language learning material in the web, as well as in digital bodies of texts. The resulting text samples correspond to the students’ individual learning levels as well as to topics of interest. Thus, it is possible to search an authentic text, which provides exactly those grammatical phenomena (such as subordinate or passive clauses) which are dealt with in class. Teachers will receive learning materials which correspond to the students’ skills on a linguistic and contentual level. Therefore, this search engine is also most suitable for teachers of integration courses for German as a second language.

In an accompanying study, the Mercator Institute develops evaluation criteria for the texts identified by the search engine and then examines the texts researched by KANSAS with regard to their didactic subject suitability for teaching.