Danish Sermons

This repository presents a corpus of 11,955 sermons from the Evangelical Lutheran Church in Denmark (ELCD), which is structured for a relational database. The repository only includes metadata, since the corpus texts contain sensitive personal information.

Metadata

The corpus metadata is composed to mask the identities of individual pastors. The data is included in the file metadata.csv as a table with the following columns. The labels are composed in Danish.

```

Column name | Description

dokumentId | id of the document holding the sermon's text præstId | id of the pastor who wrote the sermon dato | date of the church service helligdag | church holiday sogn str. | inhabitants in the parish for which the sermon is composed (ELCD members/total) stift | name of diocese årgang | the pastor's birth year køn | the pastor's gender uddannelsessted | the pastor's place of education

```

Data Model

The data has been normalised to avoid ambivalence. In order to resemble the relationship between data points, the columns in the table have been divided between two distinct data models: Prædiken (sermon) og Præst (pastor). The models each describe a data class. The classes and their properties are explicated in the following:

Sermon

The class Prædiken holds data points about the context for which the sermon is composed. Præst refers to the Præst who composed the sermon. Dato states the date of the church service for which the sermon was composed. Helligdag holds the name of the church holiday for which the sermon was composed. Stift mentions the diocese in which the sermon was delivered. SognIndbyggere states the total number of inhibitants in the parish where the sermon was delivered, and SognMedlemmer states the number of ELCD-members the parish has. Text contains the sermon in full text.

Prædiken: { præst: <id Præst>, dato: String, helligdag: String, stift: String, sognIndbyggere: Integer, sognMedlemmer: Integer, text: String, }

Pastor

The class Præst specifies general data about the author of the sermon. Køn classifies the gender of the pastor (1 is male, 2 is female). Uddannelsessted mentions the city in which the pastor received his or her theological degree. Årgang includes the birth year of the pastor. Præst: { køn: Integer, uddannelsesSted: String, årgang: Integer, }

Data Graph

The models and their relationhsip is shown in the graph below. Integers are represented in red, strings in green and the data classes hold separate colour codes.