On Aggregated Fuzzy Relations and Fuzzy Sets in Computer-Assisted Assessment
Sprache des Vortragstitels:
Englisch
Original Kurzfassung:
Computer-assisted assessment (CAA) comprises any process of assessing knowledge, opinions and/or skills with the help of a computer whereas the involvement of the computer can be accomplished in very different ways. Most often such CAAs are carried out to find (at least partial) answers to questions like: Does the student know sufficiently enough about a topic to pass? Is the answer to a concrete question correct? Does it solve the question completely? How similar are presented answers? Does the student's answering behavior resemble a typical pattern?
As indicated, a couple of these questions can be answered just by ''yes'' or ''no'' but some of them are a matter of degree. Moreover, several reflect evaluation strategies following several sub-criteria such that partial results have to be summarized into a global evaluation.
We will exemplify with typical questions from CAA systems how fuzzy sets and especially fuzzy relations can be applied for modelling evaluation strategies on the level of sub-criteria. Further, we will turn to the aggregation of such relations with the purpose to get new fuzzy sets or fuzzy relations with specific properties, e.g., for modelling the similarity of two students' results at some exam. As such we try to answer some of the preceding questions. Moreover, we will discuss further application aspects of the such studied approaches to fields like data base queries and prototype based data analysis.
Sprache der Kurzfassung:
Englisch
Vortragstyp:
Eingeladener Vortrag an Universität
Vortragsdatum:
20.06.2005
Vortragsort:
Vereinigte Staaten
Details zum Vortragsort:
BISC, University of California, Berkeley Berkeley, California
Forschungs-