What to do with The Global Jukebox "data" and "codings" files?

[linneakirby]

Each of the eight database repositories in The Global Jukebox contain a "data.csv" and a "codings.csv" file. These files contain information specific to The Global Jukebox and which cannot be reconciled. However, this information could still be beneficial to anyone using LinkedMusic. What should we do with these files?

An example codings file:

An example data file:

[liampond]

This is an instance where it could be nice to be able to create our own properties and entities with something like Wikibase. I don't think we can do anything with our current workflow, but it's definitely an issue to consider because this data is valuable.

[fujinaga]

What do data.csv and codings.csv represent?

[linneakirby]

They represent the research data compiled by The Global Jukebox regarding each song found in the full database. The codings were created by The Global Jukebox team. There is a corresponding paper found here: https://osf.io/preprints/psyarxiv/4z97j_v1

[fujinaga]

Can you summarize what the contents in these two files represent?

[linneakirby]

For each of the eight datasets, the data file lists each performance/case and provides the properties and values associated with them (for example: V1 has a value of 2). The codings file provides a key to interpret these values.

For example, in the Parlametrics dataset, V1 is "Overall pattern of interaction" and a value of 2 means that within the conversation, the interaction is split between speakers A and B such that there is a "Predominance of A with B's contribution confined to some interjections".

[fujinaga]

Thank you. I think we can leave these things out of our data lake for now.