Can I get text XY as an e-text, please? 

Of course! Under “Works” in the work view, you can download the texts either individually or as a package as \*.txt. Here are detailed instructions. 

Technical instructions for using the MHDBDB: Reading texts

Step 1: Search for a work

1. Go to the website for the works in the MHDBDB: [List of works](https://mhdbdb.sbg.ac.at/#/work/list).

2. Use the search function to find the desired work. 

Example: [Iwein](https://mhdbdb.sbg.ac.at/#/work/view/work_602). Make a note of the abbreviation (IW).

Step 2: Wait for the metadata to be generated

- After selecting a work, the metadata is calculated in the background. This may take some time.

Step 3: Accessing the text

- Read online: Click on the scroll icon to read the text online.

- Download (in the future): Click on the green icon as (function becomes available 2026).

- Current status: The green icon indicates whether the text is available for download and displays the text's sigle (e.g., IW for Iwein).

Alternative: Text retrieval via GitHub

1. Open the MHDBDB GitHub account: [MHDBDB GitHub](https://github.com/Middle-High-German-Conceptual-Database).

2. Choose between the categories TEI texts or plain txt texts, depending on your needs.

3. Download the file with the corresponding sigle (e.g., IW) or the entire corpus.

How do I cite the MHDBDB? 

Middle High German Term Database (MHDBDB). University of Salzburg. Coordination: Katharina Zeppezauer-Wachauer. Since 1992. URL: [http://www.mhdbdb.plus.ac.at/](http://www.mhdbdb.plus.ac.at/) (Date accessed). DOI: 10.60646/MHDBDB. 

Can I save my search queries?

It is already possible to view the search history for a session. A personalized user area was planned, but will probably no longer be implemented.

Which Metadata are available?

The MHDBDB uses BIBFRAME 2.0 (RDF) for bibliographic metadata and CIDOC-CRM (RDF) for other metadata objects (people, places, events). All metadata in the MHDBDB is linked to the Semantic Web and standard data and made available as Linked Open Data (LOD). 

Why are the editions so old?

This is due to copyright reasons. However, in this age of digital editions, we are keen to implement new open access editions, provided they are available in TEI format. Please contact us if you have any edition that you want to integrate! 

Why does the search take so long?

The MHDBDB is based on RDF (graph data) in the background and is a huge data battleship. By way of comparison, conventional DH projects have an average of hundreds of thousands to a maximum of a few million [triples, (https://www.digitale-edition.at/o:konde. 167)], i.e., graph statements, in the single-digit range. Larger ones (e.g., Europeana) have 300–500 million. Due to its size, the MHDBDB has 6 billion triples!

Data processing requires much more power than pure data storage. The number of triples depends on various factors.

- Data volume: The more data sources and information are included in the project, the more triples are created. 

- Complexity of the models: If the RDF model is complex and represents many relationships between entities, this leads to a higher number of triples. 

The MHDBDB has been in use for over 50 years and has 10 million tokens. Each of these tokens refers to 10-20 different annotations (cf. data), which in turn are linked to each other in numerous complex relationships. 

Some queries can currently take up to 40 seconds. Perfect for grabbing a coffee! 

Will this be faster in the future? 

We are continuously working on optimizing our search capabilities. However, graph data eventually reaches its limits with such complex models and large amounts of data. The MHDBDB uses GraphDB, an established RDF triplestore that is optimized for storing and querying large semantic graphs. Nevertheless, it is not sufficient for completely satisfactory performance. 

Ultimately, we are dependent on decisions and developments at the University of Salzburg, as there is currently no triplestore/repository known to us, either within or outside universities, that can easily handle the large volumes of the MHDBDB. We are also working with other partners on a better infrastructure solution in the high-performance computing (HPC) sector for Austrian digital humanities projects [see [www.dhinfra.at](http://www.dhinfra.at)]. 

Until such a “supercomputer” for DH projects is available, we ask for your continued patience.

Why should I use the new platform even though it doesn't work perfectly yet? 

There are a large number of new features, advantages, and improvements compared to the old website, such as the ability to download full texts and annotations, a visual search mask for easy single-word searches, a connection to the Semantic Web, and much more. We are aware that the new MHDBDB is still under development and aim to fix persisting problems.

How long will the “old” MHDBDB continue to exist?

Unfortunately, we will be forced to take the “old” MHDBDB offline at the end of 2026. This is due to serious security vulnerabilities in the server architecture, which the IT department at the University of Salzburg is no longer willing to tolerate. However, we are working on a new solution to bring you all the features in a new context.