ECOLM - An Electronic Corpus of Lute Music
The database
Clusters
Clusters are the result of an attempt to address various problems of how historical objects may
be variously grouped together by musicologists and musicians. Almost any data type may be classed
as belonging to any number of clusters - clusters can even contain other clusters. The purpose of
a cluster is given by its type as drawn from Cluster_Types. These include the vague
grouping of ‘project’ for entities that are relevant to a study, various concordance
relationships and computer-generated groupings. The main Clusters table is set
up as follows:
| Cluster ID | integer | Auto-increment-generated primary key |
| * ClusterType ID | integer | Joins with Cluster_Type table, giving type of cluster (see below) |
| Cluster Name | string | Arbitrary, but preferably unique, intuitive and memorable |
| Cluster Description | text | Unstructured text explains reason for defining cluster. |
| Cluster Owner | integer | Identifies any group with write permissions. 0 means writeable by any contributor.
Otherwise refers to Group ID from Groups table. |
| Cluster Owner | integer | Identifies any group with visibility permissions. 0 means visible to any user.
Otherwise refers to Group ID from Groups table. |
| Cluster Created | date | Creation date |
| Cluster Modified | date | date of most recent modification |
The look-up table Cluster_Types is simply:
| ClusterType ID | integer | Auto-increment-generated primary key |
| ClusterType Name | String | Short identifier |
| ClusterType Description | String | Longer explanation |
It is intended that historical events should be attachable to clusters, but this feature has not yet been implemented.