All About Data – Exploratory Data Modelling and Practical Database Access

One of the main challenges in Digital Humanities is to determine which data actually is relevant to a given research question, and how this data can be stored, analysed and processed. Modern database technologies provide technical solutions to data storage and access, but they require that data be organised in specific ways. The workshop will focus on identifying the data objects of interest, on deriving a database schema from these objects, and on querying relational or graph database systems.

Relational databases organize their data in flat tables. SQL is the standard query language to search for and extract data from the database. The technology is mature, there are many excellent database systems available, and most programming languages and application programs provide easy access to relational databases.

Graph databases organize data objects as unique information nodes linked via named arcs. Cypher is a popular query language for graph databases; its queries are compact but easy to read. In graph databases, the database schema evolves with the data inserted into the database. This makes graph databases ideal for exploratory data modelling.

In the workshop the technologies will be introduced with sample data from a digital humanities field, e.g. discourse analysis. In hands-on exercises, participants will work on their own data, or with the sample data provided. We will discuss ways of asking questions to the data and then try to express them in the query languages.

The first week will cover data modelling in graph databases. Given a collection of documents, we will identify the objects of interest. Typical questions we have to answer concern the objects' required or optional properties, whether we can we find common features, and what type of relationships exist between the objects. We will formulate queries using the visualisation provided by the graphical user interface.

The second week will cover relational databases. We will develop an entity relationship model of our data and systematically convert this into a relational database schema. We will then learn to formulate SQL queries, including aggregate functions. Finally, we will have a look at accessing relational databases from external applications, e.g. statistics packages or programming languages.