TechNical Specifications & Documentation
Here, we share the documentation about our process and the technical infrastructure for the portal, MAAS1848.umn.edu. The software ETL code can be found on GitHub and the Data Repository of the University of Minnesota (DRUM) declares attribution for our open-source code.
Front End Infrastructure: The MAAS1848 portal uses Omeka S, an open source software for curating art materials, to provide a front end user interface that is easy to navigate and update. Omeka S has been a proven software in the arts industry to showcase collections and individual works of art. The Omeka S software has built-in database functions and plug-ins that allow our administrators to quickly build the site using simple widgets and tools without writing any code.
Backend Infrastructure: For institutions with an existing application programming interfaces (API), our team built Extraction-Transformation-Loading (ETL) software using Python 3, a standard language, for information transfer into MAAS1848. For partners without an existing API, we received a .csv file for ingestion. This strategy is ideal for the portal's second-iteration and collaborations with underfunded cultural centers.
The ETL scripts are stored on GitHub and with the University of Minnesota to ensure longevity and sustainability for the code.
Hardware: The MAAS1848 application is hosted on a set of servers provided by the University of Minnesota Office of Information Technology (OIT), and managed by LATIS (Liberal Arts Technologies and Innovation Services). The application has dedicated development and production servers which are deployed in a VMWare cluster. These are each configured via the Ansible configuration management platform. The application leverages a MySQL database hotel provided by OIT, with dedicated databases for the development and production instances.
Data Management Plan: The development and production instances, along with their databases, are mirrored across multiple physical locations and have fourteen days of backup retention. OIT is responsible for security patching and monitoring, and there are system-level converges managed by Chef every 30 minutes to ensure the environment matches the approved and audited baseline. LATIS staff are responsible for the application environment, including the creation of the Ansible roles, which are archived in GitHub.
The data management plan includes detailed technical documentation on how to manage, operate, and deploy the Omeka S instance on the University of Minnesota’s servers as well pre-recorded training videos and hands-on training for the team.
Umbra Search Technical Review.pdfWhy did we select Omeka S for our portal?
The MAAS1848 portal uses Omeka S, an open source software for curating art materials, to provide a front end user interface that is easy to navigate and update. Omeka S has been a proven software in the arts industry to showcase collections and individual works of art. The Omeka S software has built-in database functions and plug-ins that allow our administrators to quickly build the site using simple widgets and tools without writing any code.
II. Institution Data Extraction Exercises--PUBLIC.pdfHow did we select records from targeted libraries and museums?
This document outlines the data extraction exercises performed by the co-PIs (Karen Mary Davalos and Constance Cortez), Data Curator (Mary Thomas), community stakeholders, and technical staff to identify potential records to be imported into the MAAS1848 Portal. This process included conducting assessments of the contributors, collections, and institutions that could be culled for the first iteration of the MAAS1848 Portal.
Co-PIs' expert knowledge guided our extraction exercises for Calisphere, Digital Public Library of America (DPLA) and The Portal to Texas History and the Smithsonian American Art Museum. A different strategy, as noted, was used for the International Center for Arts of the Americas (ICAA). For each of these aggregators, our team identified relevant collections, records, or content for MAAS1848 and provided these suggestions to the technical team to pull the appropriate content from each aggregator.
IV. Rhizomes Schema.pdfWhat is the metadata schema for our portal?
The MAAS1848 portal harvests data from a variety of sources, each with their own metadata scheme. This document shows how we map their metadata to our own schema.