This page contains all of the information you need to set up your own Memory Bank.
We would be REALLY keen to hear how you are using the resources on this page and the memory bank itself.
Please email Sian to schedule a quick teams call to tell us about your experience and to answer any questions you might have: s.e.lucas@stir.ac.uk
Video Guidance
Malte has created a series of videos to talk you through the Memory Bank and how to use it. Please click on the links below to watch:
https://myshare.uni-osnabrueck.de/f/3194375454e6415a92af
https://myshare.uni-osnabrueck.de/f/3194375454e6415a92af/?dl=1
https://myshare.uni-osnabrueck.de/f/3194375454e6415a92af/?dl=1
Memory Bank Guides
We have also created user guides that should help you make sense of what the Memory Bank is and how to use it. The guides also have some questions for you to think about in relation to what you might want to use the Memory Bank for, who should make decisions about what materials are uploaded, who uploads them and who has access. There are no hard and fast rules around this. Its for you to decide.
The Memory Bank Quick Start Guide
The Memory Bank Guide for Staff
This guide includes some additional information which helps you think through how the Memory Bank complies with Data Protection regulations. We have included a draft Data Protection Agreement which might help you with this. Please be aware that Data Protection requirements might change. You can get up to date information here For organisations | ICO
Memory Bank Software
The software was co-designed by children, young people and adults in Germany and Scotland who worked closely with Malte to produce the look and the functionality of the Memory Bank. The software is open source which means it is free to download and install.
You can download the Memory Bank by clicking here
The source code is available on a GitHub repository, along with further information about the technical requirements:
https://github.com/virtuos/ARCH
Technical Overview
The Software is an open-source Python web-application based on the Django framework. We provide a Docker Image and Docker configuration files to deploy it via a Docker container. As a database, it is using a PostgreSQL database. As part of the research, we hosted three instances of the platform on a test server, which was running on Ubuntu 22.04.2 LTS with 12 GB RAM. We were using (optional) AI models, which required additional RAM, but these features are optional and the required memory also depends on how many workers you need. For hosting a single instance, 8 GB or even 4 GB are sufficient, even though we would recommend you go for more, if possible. Scaling the service can be done, e.g. by using Gunicorn as an application server and adding more workers. The required storage depends highly on the usage and the amount of data which will be uploaded, this can also be added dynamically.