Experimental Study On The Promotion Of Deep Learning By The Representation Of Language Of Multimedia Design

Horizon report pointed out that the future education would turn to deep learning.Related organizations invested a lot of resources to supply the best quality education resources to the learners.These resources were particularly important to the learners.In order to make these education resources more attractive,to achieve better learning effect,researchers focused on how to design resources and media for these education projects.In fact,the existing form of education had gone beyond the traditional teaching environment,The new teaching activities and learning methods paid more attention to participation,such as MOOCs.It was believed that the well-designed curriculum resources would enhance the learning process and results of the learners.However,the new learning methods also brought the information fragmentation,needed to improve the deep learning methods,to avoid surface learning due to the fragmentation of knowledge.Research had indicated that with the following characteristics of the learning process was defined as surface learning:focusing on remembering,recalling without purpose,completing the task without active efforts,confining to the internal structure of knowledge,lack of strategy and method,jumping to conclusions without relevant thinking,concentrating on grade of coursework or desiring to pass the course rather than comprehending the learning materials,depending on extrinsic motivation rather than intrinsic motivation,failing to solve new problems independently.The main value of knowledge existed in the process of knowledge transfer and problem solving.When knowledge was used to solve new problems,the real role of knowledge could be played.In order to enhance the deep learning,focused on the development of learners' thinking was the practical path to cultivate talents.The definition of deep learning in this research was:'Learners can integrate new knowledge actively and critically,the starting point is deep understanding,the orientation is the migration in new scenario,and the target is to resolve the complex problem and to cultivate innovative ability.' It could be an evaluation criterion of the gap between the state of knowledge in the cognitive structure of learners and the original state of knowledge.But it was different with the meaningful learning,the starting point definition of the latter was based on memory of association between the old and new knowledge.The starting point definition of deep learning was based on understanding.It was deep learning to study and apply for what was higher than the original state of knowledge on the basis of understanding.Deep learning became an important research field to be explored because of following three reasons.Firstly,researchers paid more attention to deep learning in the era of big data;secondly,currently surface learning flooded various fields,from the classroom to the online learning,from traditional media to the mobile terminal;thirdly,research data of deep learning promotion was deficient.By Delphi investigation and Analytic Hierarchy Process(AHP),this research analyzed the core elements to promote deep learning:target presentation,behavior input,learning situation and reflection.The generation and promotion of the core elements provided guidelines and key points for realization of deep learning,and made significant sense of the popularization of deep learning in educational practice.This research also focused on how to characterize the core elements of deep learning by Multimedia Language.The specific presentation of the target is the connection of the ideal learning methods and the actual situation of learners,from the educational ideal,and based on the actual situation of the learners.The presentation of the target not only affected the effect of the actual classroom teaching,but also affected other educational methods.Multimedia Language representation of the target made significant sense of the migration and problem solving.The aim of this research was to investigate objectives which were represented by Multimedia Language in different levels influenced on deep learning.And knowledge types,the effects of declarative knowledge and procedural knowledge needed to be measured.Behavior involvement is one of the most important input factors in the learning process,which directly affected deep learning effect.Although many researches had been carried out in the past,such as behavioral or cognitive input,they hadn't set deep learning as the logical starting point.These researches assumed that when learners solved challenging problem with and paid high attention,it can made clear their behavior input.But in fact they might not,the cognitive load theory explained the extraneous cognitive load was also likely to lead to a large part of the behavior of the input was meaningless.The experimental research module is studying the core elements of deep learning,which included two studies,twelve eye movement experiments and twelve cognitive behavior experiments in this research,and the subjects were from freshman,sophomore and junior.Experiment ?,the eye movement experiments and cognitive behavior experiments studied first fixation time,total fixation time,total fixation count and learning effects of these subjects which learned with different level of objectives.Experiment ?,there was also two experiments about Multimedia Language representation of behavioral engagement.The eye movement experiments and cognitive behavior experiments studied the first fixation time,total fixation time,total fixation count and learning effects of these subjects which learned with different levels of behavioral engagement.Based on the results of cognitive behavior experiments,we analyzed from the following dimensions:? the comprehension scores of new knowledge?the scores of internal related transfer?the scores of external extended transfer.In this research,creative achievements were as:? Re-defined the concept of deep learning through analyses as well as combining teaching practice and experiment on deep learning.? Worked out the element model on deep learning through combing quality research and quantity research;provided the basis of practicing deep learning effectively.? Presented the process model of deep learning and provided executable action guideline to establish a complete deep learning process.? Established the evaluation model of deep learning effect.Improved former retention and migration evaluation method to refine internal related transfer and external extended transfer based on new knowledge comprehension.? Established the corresponding relationship between the experimental study of eye movement and the effect of deep learning.? Presented a series of learning elements and product rules of representation of Multimedia Language,which provided the basis for the design and development of the media to achieve deep learning standards.