HASKI Backend System

This is the offical Repository for the HASKI (Hochschullehre: Adaptiv, selbstgesteuert, KI-gestützt) Backend. It will provide all necessary tools for creating and recommending learning paths and learning elements for the students, who will use the system. This is a joint project from three bavarian universities:

Ostbayerische Technische Hochschule Regensburg

Technische Hochschule Aschaffenburg

Hochschule Kempten

The HASKI-Project is supported by the German Federal Ministry of Education and Research (BMBF), grant number 16DHBKI036, and the Bavarian State Ministry of Scie The responsibility for the content of this publication remains with the authors.

Purpose of the Project

The overall HASKI concept is designed to provide learners with adaptive self-directed learning. These are the learning space, the adapted Learning Management System (LMS) and the HASKI system. The actors are the learners and teachers. The adaptive learner model adapts to the learner and not vice versa. The domain model (documents, videos, podcasts, learning tasks, ...) is refined so that the tutorial model appropriately accommodates students with their different expectations and prior knowledge with tailored learning paths and feedback. Adaptive learning and testing means that learners are presented with tasks that match their learning and skill level. This is a condition of success for adaptive teaching and learning arrangements. The overall HASKI concept is based on a blended learning approach. Students receive learning materials via an LMS in which learning paths are created individually according to learner, tutor and domain models. In this process, learners can determine how their learning process looks like depending on which learning path they choose. All actions of the learners, as well as the solutions of the learning tasks and quizzes are transmitted to the HASKI system - an AI-supported software. The HASKI system determines feedback from this, which is passed directly to the learners via the LMS. The AI system also customizes learning paths for learners based on this data. Furthermore, the HASKI system creates a report of the learning activities of the learners for the teachers.

Technology

The Backend System is developed as a Python Web-App with the Flask Web Framework. It will provide RESTful-API Endpoints for the communication. The overall system follwos the principle of loose coupling and a layer-architecture. The three models (student, domain, tutoring) are all seperated as an extra domain and therefore are loosely coupled. The system will work with the Service-Layer-Pattern, the Unit-of-Work-Pattern and the Repository-Pattern.

Development

We will follow some Gudielines for the Development. These are:

• We will use Python 3.10 for our development.

• We will follow the Zen of Python.

• We will write idiomatic python code and follow the PEP8 style guide.

• For better code, we will use the library pycodestyle for linting.

• We will use Flask for the overall creation of the backend.

• We will create RESTful-APIs and document them with OAS 3.0.

• We will use the pytest library for our TDD.

• We will use TDD for all parts. No tests, no merge.

• The used libraries will be saved in a requirements.txt file, so that everyone will be able to directly install the necessary parts.

• Uses as many comments as necessary, as few as possible.

• The naming (of variables, functions, classes, etc.) should be self-explaining.

• We use logging for debugging and error handling: logger.debug, logger.info, logger.warning, logger.error, logger.critical

Setup

Crypto

In order for the system to verify requests, JWTs with RS256 signed signature will be used. Therefore create two files which hold public and private key in PEM format:

• keys/public.pem
• keys/private.pem

You can put them in a custom directory, but add the env varibale: JWT_KEYS_LOCATION with the path to the directory. If not set, the default is the keys directory in the root of the project. Make sure you have OpenSSL installed. You can use the following commands to generate the keys:

openssl genrsa -out keys/private.pem 2048
openssl rsa -in keys/private.pem -outform PEM -pubout -out keys/public.pem

Also remember to change the cryptographic key in ToolConfigJson.py to match yours from moodle. The URL to obtain the key is also in the ToolConfigJson.py file.

Project

Either create a new conda environment or use venv. The following steps are for conda:

• Create a new conda environment using conda create --name HASKI-Backend --file requirements.txt
• Activate the environment: conda activate HASKI-Backend

Using venv:

• Create a new venv using python -m venv venv
• Activate the environment: source venv/bin/activate or venv\Scripts\activate on Windows
• Install the requirements using pip install -r requirements.txt

Then:

• Fill the .flaskenv file with the necessary information. For this, use the .flaskenv_template file as a template. Note that DB_PASSWORD is the password for the database user, not pgAdmin.

  NOTE: LTI_ACTIVITY_ID is the id of the activity in moodle. You can find it in the url of the activity. For example: https://moodle.haski.de/mod/lti/view.php?id=2 CLIENT_ID is obtained in the setup of the LTI tool in moodle. It is the consumer key. KEY_ID is the key id of the key pair used for signing the JWTs. It is obtained when opening the URL /mod/lti/certs.php KEY_N_VALUE also optained in the certs page.

• Create the database as explained in DB Setup

• Run the app using flask run

To create a new requirements file after installing a new library, please run the following command: pip list --format=freeze > requirements.txt

DB Setup

For creating a (local) setup for the PostgreSQL Database for the HASKI project, you need to install PostgreSQL. You can follow these tutorials:

https://www.postgresqltutorial.com/postgresql-getting-started/install-postgresql/

https://www.postgresqltutorial.com/postgresql-getting-started/connect-to-postgresql-database/

After the installation, you can run it on the CMD or the pgAdmin 4 application. You first have to create a database by using python db_setup.py. Then run python db_fill.py to fill the database with mock data if applicable. Run python db_clean_up.py to clean up the database if necessary. Use arg --dropdatabase to drop the database.

Running in Docker

This project includes a Pipeline, that will automatically create a Docker image in the repository when merging into the main branch. This Docker image should always be the latest running version. Please create a .flaskenv file first locally, so that it can be passed with docker run. A template is provided in the .flaskenv_template file in this repository. Change the values as needed. Use the volume flag to mount the keys folder into the container. Start the Docker image as container with the following command: docker run docker run -v /yourfolder/keys:/app/keys -d -p 5000:5000 --env-file <path/to/.flaskenv-File> ghcr.io/haski-rak/haski-backend:main

Contribution

For contributing to the project, please work on the Issues and fulfill the requested tasks. Before commiting, always run python validate_script.py in the root folder and check the output. The test coverage should be as good as possible and there shouldn't be any information displayed by flake8, isort and black. For checking the code quality, you can use Sonarqube (https://docs.sonarqube.org/latest/setup-and-upgrade/install-the-server/). There shouldn't be any code smells left, when contributing to the project.

After a Pull Request, one of the Code Owners will check and give feedback before the merge is possible.

Moodle

The functionality to send requests to moodle via REST-API needs to be activated in moodle settings. The following settings are necessary, do not forget so save the changes:

• Active web service protocols: REST protocol (enable)
• Enable web services: yes
• Enable web services for mobile devices: yes

Also it is necessary to create a token for a user with admin rights.

• Go to Site administration -> Server -> Manage tokens -> Create token
• Select an admin user and create a token with the Service "Moodle mobile web service", set the validity to your needs and save the token.

Lastly you have to add the following lines to your .flaskenv file:

• REST_LMS_URL= "yourmoodleurl" (something like https://moodle.haski.de)
• REST_TOKEN="yourtoken" (something like f5b78d90cff0bf8f61a5745ce1441e87)