摘 要：“ACT-R”是一种类似OAR 架构的网状结构认知模式，又叫做推理思维的自适应控制模型，它能在储存知识时按照情境的“模块”进行组装。模块化方式能帮助学习者快速分析问题，增加知识的数量，评价核心问题，提升认知深度，创新问题结构，促进知识迁移。在“导学激趣”ACT-R 模式上运用深度学习CNN 算法，使教者苦教与受教者苦学的状况得以改变。目前，深度学习算法运用比较成功的是CNN 卷积神经网络算法，在“导学激趣”ACT-R 模式过程中引入CNN，能在激发学习者学习兴趣的同时，增强学习的自信心，有助于促进学习的深入，学习者的思维也能得到启迪，潜能得到开发，有助于学习者发展自身的个性品质。本项目基于E-Learning 时代背景，从深度学习的角度出发，探索研究“导学激趣”ACT-R 模式的基本思路，帮助学习者捕获感知、行为和思想，组织生活语言情景产生智能行为，用以仿真并理解人的认知。
关键词：导学激趣；深度学习；E-Learning；ACT-R 模式
中图分类号：TP18 文献标识码：A 文章编号：2096-4706（2018）09-0136-03
Research on the ACT-R Mode of“Guide Learning to Stimulate Interest”Based onCNN Algorithm of Deep Learning in the Context of E-Learning
HUANG Ling
（Hunan Institute of Information Technology，School of Electronic Information，Changsha 410151，China）
Abstract：“ACT-R”is a network structure cognitive model of SOAR architecture，also called an adaptive control model of reasoning thinking，which can be assembled according to the“module”of the situation in the storage of knowledge. Modularization can help learners to analyze problems quickly，increase the number of knowledge，evaluate the core issues，improve the depth of cognition，innovate the structure of the problem，and promote knowledge migration. In the“guided learning interesting”ACT-R mode，weuse the deep learning CNN algorithm to change the state of hard teaching. At present，the CNN convolution neural network algorithm is successfully used in the depth learning algorithm. The introduction of CNN in the“guided learning interesting”ACT-R model can stimulate learners’ interest in learning，enhance their confidence in learning，and contribute to the deepening of learning. The learners’ thinking dimension can also be enlightened and potential development. It is helpful for the learners to develop their own personality traits.This project is based on the background of the E-Learning era. From the perspective of deep learning，this project explores the basic ideasof the“guided learning interesting”ACT-R model，helping the learners to capture perception，behavior and thought，and organize life language scenarios to generate intelligent behavior to simulate and understand people’s recognition.
Keywords：learning to stimulate interest；deep learning；E-Learning；ACT-R mode
参考文献：
[1] 王天欣，张征. 基于ACT-R 模型的互联网交互模式库设计 [J]. 计算机与数字工程，2014，42（6）：1095-1099+1103.
[2] 邓渝.E-learning 背景下在线学习模式在高校教学中的应用研究 [J]. 教育教学论坛，2018（20）：78-79.
作者简介：黄玲（1984.10-），女，湖南岳阳人，讲师，硕士，研究方向：计算机技术、图像处理。