AIEMS项目 (AIEMS Project)

Project Title: “AIEMS( Artificial Intelligent Education Monitoring System): Development of an Advanced Artificial Intelligent (AI) Monitoring and Feedback System for Online Education ”

项目名称：“ AIEMS(人工智能教育监控系统)：开发用于在线教育的高级人工智能(AI)监控和反馈系统”

第2章 (Chapter -2)

人工神经网络架构 (The ANN Architecture)

In our previous chapter we focused on the concept of a SMART classroom with embedded artificial cognitive ability. Due to the pandemic still in effect and classes shifting mostly online, the architecture of a SMART classroom should offer innovation and address challenges of the traditional education platform.

在上一章中，我们重点介绍了具有嵌入式人工认知能力的SMART教室的概念。 由于大流行仍然有效，班级大部分都在网上转移，因此SMART教室的架构应提供创新并应对传统教育平台的挑战。

We had highlighted two major challenges of online education (i) enhancing on line proctoring mechanism that ensures academic honesty (ii) providing a timely and cost-effective feedback mechanism on the performance of students.

我们强调了在线教育的两个主要挑战(i)增强在线督导机制，以确保学术诚信(ii)为学生的表现提供及时且具有成本效益的反馈机制。

Another major challenge as an educator is identifying the learning style of students. Most students can be easily categorized into the following sections:

作为一名教育者，另一个主要挑战是确定学生的学习方式。 大多数学生可以很容易地分为以下几部分：

(1) Verbal: They prefer to learn through keywords.

(1)言语：他们更喜欢通过关键词学习。

(2)Visual: They prefer to learn through pictures.

(2)视觉：他们更喜欢通过图片学习。

(3) Auditory: They prefer to learn through sound/rhythm.

(3)听觉：他们喜欢通过声音/节奏学习。

(4) Physical/Kin-aesthetic: They prefer to learn by physical touch.

(4)身体/肌理：他们更喜欢通过身体接触来学习。

(5) Logical/Mathematical: They prefer to learn through reasoning.

(5)逻辑/数学：他们更喜欢通过推理学习。

As an educator, it is not always possible to address all the different learning styles in the lesson plan. Artificial Intelligence (AI) technology can transform this challenge into reality and offer more flexibility. This chapter primarily focuses on the Artificial Neural Network (ANN) architecture that has been used in developing our project AIEMS. The motivation for AIEMS is developing a SMART virtual classroom that can offer a holistic learning experience to its student members. Simultaneously it will address the challenges of academic dishonesty by AI-enhanced proctoring mechanism.

作为一名教育者，并非总是可以在课程计划中解决所有不同的学习方式。 人工智能(AI)技术可以将这一挑战变为现实，并提供更大的灵活性。 本章主要关注人工神经网络(ANN)架构，该架构已用于开发我们的AIEMS项目。 的 AIEMS的动机是开发SMART虚拟教室，该教室可以为其学生成员提供全面的学习体验。 同时，它将通过AI增强的监督机制来应对学术不诚实的挑战。

Deep Learning (DL) has been used in this particular model application for developing AIEMS platform, as it can be easily applied to a multi-layer neural network. An artificial neural network (ANN) is a mathematical model which is inspired by the architecture of a neuron (the basic working unit of our brain). The basic architecture of an ANN is shown below:

深度学习(DL)已在此特定模型应用程序中用于开发AIEMS平台，因为它可以轻松地应用于多层神经网络。 人工神经网络(ANN)是一种数学模型，其灵感来自神经元(我们大脑的基本工作单位 )的体系结构。 ANN的基本架构如下所示：

Mathematically, a single layer neural network can be represented by the following equation (1)

在数学上，单层神经网络可以由以下等式(1)表示

The learning analytics require the processing of big data. Therefore a self-adaptive computational algorithm can allow for effective analysis through curating patterns and simultaneously accumulating new input data into this model. Deep learning (DL) has shown application and scope in the fintech industry and is constantly utilized in detecting fraudulent activities.

学习分析需要处理大数据。 因此，自适应计算算法可以通过选择模式并同时将新的输入数据累积到该模型中来进行有效的分析。 深度学习(DL)已显示出在金融科技行业中的应用和范围，并不断用于检测欺诈活动。

了解平台的基础架构 (Understanding the infrastructure of the platform)

AIEMS is a virtual e-learning platform that will contribute to one of the few major purposes in a SMART classroom environment:

AIEMS是一个虚拟的电子学习平台，将有助于实现SMART教室环境中的几个主要目的之一：

Identify the cognitive learning style of the student.

识别学生的认知学习风格。

This is particularly important because, in the most classroom, the lesson delivered by a mentor or educator are usually standardized and skewed towards student audience with a particular learning style or ability. This eventually has a great impact on the development of a student’s full potential. The aid of ANN, can accurately identify the student’s inherent learning style and hence the student can be provided with the style of learning material that is ideal for their learning and development. Similarly, an educator needs to be aware of their performance by analyzing and sometimes predicting the academic and social changes occurring in the SMART classroom (or virtual classroom).

这一点特别重要，因为在大多数教室中，导师或教育者提供的课程通常是标准化的，并且倾向于具有特定学习风格或能力的学生听众。 这最终将对学生的全部潜能发展产生重大影响。 借助人工神经网络，可以准确识别学生的内在学习风格，因此可以为学生提供最适合其学习和发展的学习材料风格。 同样，教育者需要通过分析并有时预测SMART教室(或虚拟教室)中发生的学术和社会变化来意识到自己的表现。

For the ease of understanding the cognitive learning style of the student, it can be supplemented with a flowchart. The following flowchart can be used to identify if the student is a visual learner.

为了便于理解学生的认知学习方式，可以在流程图中进行补充。 以下流程图可用于识别学生是否是视觉学习者。

The attached flowchart is a simplistic representation of the model used to determine a visual learner. Since visual learners are adapted to the learning process through pictures and diagrams, therefore a lesson plan can be presented by the instructor containing pictures. After the material is presented to the student, they can be asked a series of questions related to the material. If the student can answer accurately with a score of 50% and above then the student can be considered a Visual Learner.

所附流程图是用于确定视觉学习者的模型的简化表示。 由于视觉学习者可以通过图片和图表适应学习过程，因此讲师可以通过包含图片的教学计划来提出课程计划。 在向学生展示材料后，可以向他们提出一系列与材料有关的问题。 如果学生能够以50％或更高的分数准确回答，则可以认为该学生是视觉学习者。

Though it might not be a very accurate model and there are underlying challenges. It, however, does offer an opportunity to advance the current learning platform offered to students and mentors alike.

尽管它可能不是一个非常准确的模型，但存在潜在的挑战。 但是，它确实提供了一个机会，可以改进为学生和导师提供的当前学习平台。

网络参考 (Web references)

[1]https://sites.google.com/site/holisticappeducation/engaging-different-learning-styles

[1] https://sites.google.com/site/holisticappeducation/engaging-different-learning-styles