Instructor Dashboard for E-Learning Systems using Machine Learning

Software Requirements Specification

Version 1.0

Group Id: S2402AE256 (BC210200009, BC210200101)

Supervisor Name: Mehboob Ali

 Revision History
Date Versio Description Author
n
(dd/mm/
yyyy)
20/05/2024 1.0 The project aims to develop an Instructor BC2102000101
Dashboard for E-Learning Systems using Machine BC2102000009
Learning. This dashboard will enable instructors to
track students' performance by analyzing various
data sources including student engagement,
historical records, and demographic information. By
employing machine learning algorithms, the system
will predict student performance, providing
valuable insights to instructors. The dashboard will
feature functionalities such as data exploration, heat
map plotting, performance prediction, model
visualization, and real-time model deployment,
ultimately enhancing the effectiveness of e-learning
environments.
Contents
1. Scope.....................................................................................................................................................................4
1.1 Objectives........................................................................................................................................................4
1.2 Out of Scope....................................................................................................................................................4
1.3 Constraints.......................................................................................................................................................4
1.4 Assumptions.....................................................................................................................................................4
2. Functional Requirements.......................................................................................................................................4
2.1 Data Pre-processing.........................................................................................................................................4
2.2 Model Training and Testing.............................................................................................................................5
2.3 Model Tuning...................................................................................................................................................5
2.4 Build Instructor Web App Dashboard.............................................................................................................5
2.5 Model Deployment in Real-Time Environments.............................................................................................5
3. Non-Functional Requirements...............................................................................................................................5
3.1 Performance Requirements..............................................................................................................................5
3.2 Security Requirements.....................................................................................................................................5
3.3 Reliability Requirements.................................................................................................................................5
3.4 Usability Requirements....................................................................................................................................6
3.5 Maintainability Requirements..........................................................................................................................6
4. Use Case Diagram(s):............................................................................................................................................6
...................................................................................................................................................................................6
5. Usage Scenarios:....................................................................................................................................................6
6. Adopted Methodology...........................................................................................................................................9
6.1 Overview of the VU Process Model:...............................................................................................................9
6.2 Adoption of the VU Process Model:..............................................................................................................10
6.3 Benefits of the VU Process Model:...............................................................................................................10
6.4 Conclusion:....................................................................................................................................................11
7. Work plan:...........................................................................................................................................................11
 Instructor Dashboard for E-Learning Systems using Machine Learning

1. Scope
The scope of the project encompasses the development of an Instructor Dashboard for E-Learning
Systems using Machine Learning. This dashboard aims to empower instructors by providing them with
tools to track students' performance through the utilization of data encompassing student engagement,
historical records, and demographic information. The project will leverage machine learning algorithms
within the Python framework to predict student performance, offering valuable insights to instructors.

1.1 Objectives
 Develop a robust data pre-processing pipeline to clean and prepare the input data for analysis.
 Implement machine learning models for predicting student performance based on various input
features.
 Design and develop a user-friendly web application dashboard for instructors to visualize and
interpret the predicted student performance.
 Integrate the dashboard with machine learning models to provide real-time insights into student
performance.
 Deploy the dashboard and machine learning models in real-time environments such as Azure or
AWS for practical usage.
1.2 Out of Scope
 Integration with existing Learning Management Systems (LMS) beyond data retrieval.
 Development of mobile applications or additional platforms beyond the web-based dashboard.
 Advanced analytics or machine learning techniques beyond the scope of predicting student
performance.
 Curriculum planning or course content recommendations based on student performance.
1.3 Constraints
 Availability and quality of input data from the Open University Learning Analytics dataset.
 Compatibility with selected tools and technologies (Python, Scikit-learn, Pycaret, Colab, VS Code,
Azure, AWS, Github, and Streamlit).
 Compliance with relevant legal and regulatory requirements for data privacy and security.
 Resource constraints such as time, budget, and available expertise.
1.4 Assumptions
 The availability of sufficient computational resources for model training and deployment.
 Access to the necessary APIs or libraries for data retrieval and integration.
 Adequate user training and support for instructors utilizing the dashboard.
 The stability and reliability of selected tools and platforms throughout the project lifecycle.

2. Functional Requirements
2.1 Data Pre-processing
 Data Cleaning: The system shall perform data cleaning to handle missing values, outliers, and
inconsistencies in the input dataset.
 Feature Engineering: The system shall extract relevant features from the dataset and transform
them into suitable formats for model training.
 Data Integration: The system shall integrate multiple data sources, including student
engagement, historical records, and demographic information, for comprehensive analysis.
2.2 Model Training and Testing
 Model Selection: The system shall support the selection and configuration of appropriate machine
learning algorithms for predicting student performance.
 Training Pipeline: The system shall facilitate the training of machine learning models using the
pre-processed data.
 Model Evaluation: The system shall evaluate the performance of trained models using
appropriate evaluation metrics such as accuracy, precision, recall, and F1-score.
 Cross-Validation: The system shall employ cross-validation techniques to ensure the robustness
of trained models.
2.3 Model Tuning
 Hyper parameter Tuning: The system shall perform hyper parameter tuning to optimize the
performance of machine learning models.
 Grid Search: The system shall utilize grid search or similar techniques to systematically explore
the hyper parameter space and identify optimal configurations.
2.4 Build Instructor Web App Dashboard
 Dashboard Interface: The system shall provide a user-friendly web application interface for
instructors to access and interact with the dashboard.
 Visualization: The system shall generate interactive visualizations, including plots, charts, and
graphs, to present student performance data and insights.
 User Authentication: The system shall incorporate user authentication mechanisms to ensure
secure access to the dashboard.
2.5 Model Deployment in Real-Time Environments
 Deployment Pipeline: The system shall establish a deployment pipeline for deploying trained
machine learning models in real-time environments such as Azure or AWS.
 Scalability: The deployed models shall be scalable to handle varying loads and concurrent user
requests.
 Monitoring and Logging: The system shall implement monitoring and logging functionalities to
track the performance and usage of deployed models.

3. Non-Functional Requirements
3.1 Performance Requirements
 Response Time: The system shall respond to user requests within an acceptable timeframe (e.g.,
less than 2 seconds).
 Throughput: The system shall support a minimum number of concurrent users accessing the
dashboard simultaneously (e.g., 100 users).
3.2 Security Requirements
 Data Encryption: The system shall encrypt sensitive data, such as user credentials and student
performance records, to ensure confidentiality.
 Access Control: The system shall enforce role-based access control to restrict unauthorized
access to sensitive functionalities and data.
3.3 Reliability Requirements
 Availability: The system shall maintain high availability, with an uptime of at least 99.9%.
 Fault Tolerance: The system shall handle errors and failures gracefully, with mechanisms for
error recovery and fault tolerance.
3.4 Usability Requirements
 User Interface Design: The dashboard interface shall be intuitive and easy to navigate, with clear
labels and instructions for users.
 Accessibility: The system shall be accessible to users with disabilities, complying with relevant
accessibility standards (e.g., WCAG).
 3.5 Maintainability Requirements
 Modularity: The system shall be designed with modular components to facilitate future
enhancements and maintenance.
 Documentation: The system shall include comprehensive documentation covering installation,
configuration, usage, and troubleshooting procedures.

4. Use Case Diagram(s):

Use case for project is below:

5. Usage Scenarios:
Here are the usage scenarios for each of the use cases in table format:

Use Case Title Explore Dataset

Use Case ID UC01
Actions 1. Log in to the dashboard.
2. Navigate to the "Explore Dataset"
3. Select dataset for exploration.
4. Analyze dataset.
Description 1. The instructor logs in to the web application dashboard using their credentials.
2. Once logged in, the instructor navigates to the "Explore Dataset" section of the
dashboard.
3. The instructor selects the dataset they want to explore from the available options.
4. The system presents various visualizations and statistical summaries to help the
instructor analyze the selected dataset.
Alternative
-
Paths
Pre-Conditions 1.-
2.The instructor is logged in.
3. The instructor is in the "Explore Dataset" section.
4. The selected dataset is loaded.
Post-Conditions 1. The instructor is authenticated and logged in successfully.
2. The instructor can access the dataset for exploration.
3. The selected dataset is loaded for exploration.
4. The instructor gains insights into the dataset.
Author Instructor
Exceptions Invalid credentials, network connectivity issues.
 Use Case Title Predict Student Performance
Use Case ID UC02
Actions 1. Log in to the dashboard.
2. Navigate to "Predict Student
3. Configure prediction parameters.
4. Generate predictions.
Description 1. The instructor logs in to the web application dashboard using their credentials.
2. Once logged in, the instructor navigates to the "Predict Student Performance"
section of the dashboard.
3. The instructor selects the desired input features and model configuration parameters
for predicting student performance.
4. The system applies the selected machine learning model to generate predictions for
student performance.
Alternative
-
Paths
Pre-Conditions 1.-
2.The instructor is logged in.
3. The instructor is in the "Predict Student Performance" section.
4. The prediction parameters are configured.
Post-Conditions 1. The instructor is authenticated and logged in successfully.
2. The instructor can access the prediction functionality.
3. The prediction model is configured according to specifications.
4. Predictions for student performance are displayed.
Author Instructor
Exceptions Invalid credentials, network connectivity issues.

Use Case Title Visualize Performance

Use Case ID UC03
Actions 1. Log in to the dashboard.
2. Navigate to "Visualize
3. Select performance data.
4. Generate visualizations.
Description 1. The instructor logs in to the web application dashboard using their credentials.
2. Once logged in, the instructor navigates to the "Visualize Performance" section of
the dashboard.
3. The instructor selects the performance data they want to visualize, such as actual
versus predicted performance.
4. The system generates interactive plots, charts, and graphs to visualize the selected
performance data.
Alternative
-
Paths
Pre-Conditions 1.-
2.The instructor is logged in.
3. The instructor is in the "Visualize Performance" section.
4. Performance data is selected.
Post-Conditions 1. The instructor is authenticated and logged in successfully.
 2. The instructor can access visualization tools.
3. The selected data is loaded for visualization.
4. Visualizations of student performance are displayed.
Author Instructor
Exceptions Invalid credentials, network connectivity issues.

Use Case Title Deploy Model

Use Case ID UC04
Actions 1. Log in to the dashboard.
2. Navigate to "Deploy Model"
3. Select model for deployment.
4. Deploy model to Azure/AWS.
Description 1. The instructor logs in to the web application dashboard using their credentials.
2. Once logged in, the instructor navigates to the "Deploy Model" section of the
dashboard.
3. The instructor selects the trained machine learning model they want to deploy in
real-time environments.
4. The system deploys the selected model to Azure, AWS, or similar real-time
environments for practical usage.
Alternative
-
Paths
Pre-Conditions 1.-
2.The instructor is logged in.
3. The instructor is in the "Deploy Model" section.
4. The model for deployment is selected.
Post-Conditions 1. The instructor is authenticated and logged in successfully.
2. The instructor can access deployment functionality.
3. The selected model is prepared for deployment.
4. The model is successfully deployed and operational.
Author Instructor
Exceptions Invalid credentials, network connectivity issues.

Use Case Title Authenticate User

Use Case ID UC05
Actions 1. Access the dashboard URL.
2. Enter login credentials.
Description 1. The instructor accesses the web application dashboard URL using a web browser.
2. The instructor enters their username and password in the login form provided by the
system.
Alternative
-
Paths
Pre-Conditions 1.-
2. The dashboard URL is accessible.
Post-Conditions 1. The dashboard URL is accessible.
2. The instructor's credentials are authenticated.
Author Instructor
 Invalid credentials, network connectivity issues.
Exceptions
Dashboard URL is not accessible.

Use Case Title Access Dashboard

Use Case ID UC06
Actions 1. Authenticate user.
2. Display dashboard interface.
1. The system verifies the instructor's credentials to ensure authorized access to the
Description
dashboard.
2. Upon successful authentication, the system displays the web application dashboard
interface to the instructor.
Alternative
-
Paths
Pre-Conditions 1. The instructor's credentials are entered.
2. The instructor is authenticated.
Post-Conditions 1. The instructor is granted access to the dashboard interface.
2. The dashboard interface is displayed.
Author Instructor
Exceptions Invalid credentials, network connectivity issues.

6. Adopted Methodology
For this project, the adopted methodology is the VU Process Model, which combines elements of
both the waterfall and spiral models. The VU Process Model is well-suited for projects that require a
balance between structured planning and iterative development, making it an ideal choice for software
development projects with clear objectives and evolving requirements.

6.1 Overview of the VU Process Model:

1. Waterfall Model:
 The waterfall model emphasizes a sequential approach to software development, where each phase must be
completed before moving on to the next.
 Phases in the waterfall model include requirements analysis, design, implementation, testing, deployment,
and maintenance.
 The waterfall model provides a structured framework for planning and executing software projects,
ensuring clear documentation and milestone-based progress.
2. Spiral Model:
 The spiral model incorporates iterative development and risk management principles, allowing for
flexibility and adaptation to changing requirements.
 Phases in the spiral model include identification of objectives, risk analysis, prototyping, development and
testing, and evaluation.
 The spiral model emphasizes continuous refinement and iteration, with each cycle addressing identified
risks and incorporating feedback from stakeholders.

6.2 Adoption of the VU Process Model:

For this project, the VU Process Model will be adopted to leverage the strengths of both the
waterfall and spiral models:
1. Initial Planning (Waterfall Phase):
 Requirements analysis: Gather and document detailed requirements for the instructor dashboard and
machine learning components.
 Design: Create design specifications for the dashboard interface, machine learning models, and deployment
architecture.

2. Iterative Development (Spiral Phase):

 Prototyping: Develop initial prototypes of the dashboard interface and machine learning models to
validate design concepts and gather feedback from stakeholders.
 Development and Testing: Implement features incrementally, starting with core functionalities, and
conduct rigorous testing to ensure reliability and accuracy.
 Evaluation: Review and evaluate each iteration with stakeholders, incorporating feedback and making
necessary adjustments to meet evolving requirements.

3. Integration and Deployment (Waterfall Phase):

 Integration: Integrate components of the dashboard and machine learning models into a cohesive system.
 Deployment: Deploy the integrated system in real-time environments such as Azure or AWS, ensuring
scalability, reliability, and security.

4. Maintenance and Enhancement (Spiral Phase):

 Monitoring: Continuously monitor the deployed system for performance, reliability, and security issues.
 Maintenance: Address any issues or bugs discovered post-deployment, and provide ongoing support and
maintenance as needed.
 Enhancement: Incorporate new features or improvements based on user feedback and changing
requirements, following an iterative development approach.

6.3 Benefits of the VU Process Model:

 Structured Approach: The VU Process Model provides a structured framework for planning,
development, and deployment, ensuring systematic progress and clear documentation.
 Flexibility: By incorporating elements of iterative development from the spiral model, the VU Process
Model allows for flexibility and adaptation to changing requirements and stakeholder feedback.
 Risk Management: The iterative nature of the spiral phase enables proactive risk management, with each
iteration addressing identified risks and uncertainties.
 Continuous Improvement: Through iterative development and evaluation, the VU Process Model
promotes continuous improvement and refinement of the software product, leading to higher quality and
user satisfaction.
6.4 Conclusion:
The adoption of the VU Process Model for this project combines the strengths of both the waterfall
and spiral models, providing a balanced approach to software development that emphasizes both structured
planning and iterative refinement. This methodology will enable the successful development and
deployment of the instructor dashboard for e-learning systems using machine learning, meeting the
project's objectives while accommodating evolving requirements and stakeholder feedback

7. Work plan: