2008
DOI: 10.1007/978-3-540-69132-7_41
|View full text |Cite
|
Sign up to set email alerts
|

Toward Automatic Hint Generation for Logic Proof Tutoring Using Historical Student Data

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
55
0

Year Published

2010
2010
2024
2024

Publication Types

Select...
5
2
1

Relationship

1
7

Authors

Journals

citations
Cited by 100 publications
(58 citation statements)
references
References 3 publications
0
55
0
Order By: Relevance
“…In a second, logic tutor approach, dynamic hints are constructed based on the matching of current student's problem solving path with solutions from previous submitted solutions [3] [16]. In our approach, dynamic hints are generated from a single, or at most a few, instructorprovided solution(s).…”
Section: Task Relevant Dynamic Hint Generationmentioning
confidence: 99%
“…In a second, logic tutor approach, dynamic hints are constructed based on the matching of current student's problem solving path with solutions from previous submitted solutions [3] [16]. In our approach, dynamic hints are generated from a single, or at most a few, instructorprovided solution(s).…”
Section: Task Relevant Dynamic Hint Generationmentioning
confidence: 99%
“…where from each state the best action to take is the one that leads to the next state with the highest expected reward value (Barnes and Stamper 2008). The Hint Factory uses these values when a student is in a particular state to choose the next best state from which to generate a hint.…”
Section: Hint Factorymentioning
confidence: 99%
“…A last method that we mention is the work of Barnes & Stamper [17], which was applied for the well-defined domain of logic proofs. The approach of Barnes & Stamper consists of building a Markov decision process containing learner solutions for a problem.…”
Section: Canadarmtutor)mentioning
confidence: 99%
“…In summary, in addition to some specific limitations, all systems mentioned in this section have one or more of the following limitations: they (1) require defining a body of background knowledge [6,21,22,23], (2) have been demonstrated for well-defined domains [5,6,17,21,22,23], (3) rely on the strong assumption that tasks can be modeled as production rules [6,21,23], (4) do not take into account learner profiles [5,6,17,21,22,23], (5) learn knowledge that is problem specific [5,17], or (6) require demonstrators to provide extra information during demonstrations such as their intentions or labels for elements of their solutions [6,22].…”
Section: Canadarmtutor)mentioning
confidence: 99%