Direct Execution Learning Technology - Webble Technology - page 104

5 Direct Execution in Educational Patterns

Since the early days of didactics, engaged human teachers are striving hard
to distill essential ideas frequently reusable for e cient and effective teaching-
educational memes as discussed in [12].
In scientific discourse, following the early work of Christopher Alexander [1],
repeatedly occurring structures are called patterns. To be more precise, one
should the concrete phenomena that occur instances and reserve the word pat-
tern to the general some instances have in common. According to the scientific
background of scientists involved and corresponding to the traditions and to the
methodologies of pedagogy, educational patterns are usually very vague and un-
certain [25]. The meme media idea may help to encapsulate and, thus, to clarify
several patterns of didactics.
Within the present contribution, emphasis is put on those patterns which
relate to the paradigm of direct execution. Let us begin with a pattern named
challenging the target in [12]. This educational pattern is applicable to some
settings of exploratory learning where the learning target is something being
operational such as, e.g., an algorithm. Those settings appear overwhelmingly
in studies of computer science, for instance.
If the target of learning is some algorithm, challenging the target means to
systematically probe the target by means of developing challenging input data.
Just for illustration, think of the following application cases.

- There is some sorting algorithm such as bottom-up heap sort . Usually, stu-
dents are learning about the more conventional heapsort algorithm before
they are facing bottom-up heapsort.
•

Prepare input data such that bottom-up heapsort is not better than
classical heapsort.

Prepare input data such that the advantage of bottom-up heapsort over
heapsort is enormous.
- The decision tree algorithm ID3 deserves to be called one of the oldies, but
goodies of knowledge discovery and data mining.
•

•

Challenge the algorithm by some sample as small as possible which forces
ID3 to generate a comparably complex decision tree.
- For NP-hard problems, there are usually some appealing heuristic solutions.
Especially when the problem is intuitively understandable like, for instance,
the traveling salesman problem , heuristics frequently seem to be ingenious.
Choose any implementation of one of the heuristics like, for instance, the
nearest neighbor heuristics or the insertion heuristics .
• Construct some graph on which the heuristics does not work well.

• In contrast, construct a certain example graph to which the heuristics

applies perfectly.

For a playful learning experience when dealing with those problems, direct exe-
cution is essential. Learners should be released from executing operations and,
instead, be able to put much emphasis on pondering problem details.