|
|
Based on the introduced students’ evaluation, we conclude that the proposed individual exercise task idea fosters self-directed and reflective learning. A cross-modal information augmentation method is proposed to resolve this downside. Therefore, helping college students to find suitable exercises becomes a significant downside. Thus, the precise drawback addressed on this paper is find out how to recommend exercises with excessive representativeness and informativeness from a large pool of questions. Because classification and similarity comparison are two different issues, we target our problem using a distinct baseline but comparable construction general. We suggest and practice a suitable neural network architecture for the task, and present that conditioning the model’s output for a given input text on an example exercise of the envisioned exercise type, results in an increased effectiveness, compared to an instance-impartial baseline model. At the identical time, unreflected copying of tasks already solved does not foster the understanding of the subject and leads to a false self-evaluation.
Feel free to surf to my blog post :: mitolyns.net |
|
發表於