THESIS
2020
xiii, 68 pages : illustrations ; 30 cm
Abstract
Personalized recommendation system has been widely adopted in E-learning field
that is adaptive to each learners own learning pace. With full utilization of learning
behavior data, psychometric assessment models keep track of the learners proficiency on knowledge points, and then a well-designed recommendation strategy
selects a sequence of actions to meet the unique learning objective of individual.
In this dissertation, we develop two adaptive recommendation strategies under
the framework of reinforcement learning. The first strategy involved with early
stopping enjoys a time-related learning mode with the aim of maximizing the
learning efficiency. Secondly, we consider the element of curiosity as a critical
motivate for information-seeking to propose a curiosity-driven recomm...[
Read more ]
Personalized recommendation system has been widely adopted in E-learning field
that is adaptive to each learners own learning pace. With full utilization of learning
behavior data, psychometric assessment models keep track of the learners proficiency on knowledge points, and then a well-designed recommendation strategy
selects a sequence of actions to meet the unique learning objective of individual.
In this dissertation, we develop two adaptive recommendation strategies under
the framework of reinforcement learning. The first strategy involved with early
stopping enjoys a time-related learning mode with the aim of maximizing the
learning efficiency. Secondly, we consider the element of curiosity as a critical
motivate for information-seeking to propose a curiosity-driven recommendation
policy, allowing for a both rewarding and enjoyable personalized learning path.
Numeric analyses with the large continuous knowledge state space and concrete
learning scenarios are used to further demonstrate the power of the proposed
methods.
Post a Comment