Collaborate with Lund on 19th century speaker identification

[MansMeg]

At Lund they have fixed the speaker mapping during the 19th century. We should add this to our corpus.

[fredrik1984]

After meeting with Agustin it seems that they did not get far with this (or did nothing at all). Howerver, they would like to help out fixing the mapping quality for the 19th century, e.g. see #323 in the backlog (Westac issue).

[fredrik1984]

Let´s use this as an overall issue for the current work of improving the mapping algorithm (connecting correct MP to correct speech).

• @ninpnin and @BobBorges create an aggregated CSV file with the most common unknown speaker introduction in the Swerik corpus. This should include, among other things, links to speeches, time period of common unknown speaker introductions, and a column to insert the correct MP-ID.
• Agustin Goanaga and his team at Lund University will use the CSV file to add MP-IDs to common unknown speaker introductions in the 19th century. We should also schedule a meeting with them on how to do this work.
• Lotta and Mattias can help out adding MP-IDs for common unknown speaker introductions from the 20th century + difficult cases that Agustin et al. could not fix. See also url in <pb> elem points to wrong page #24
• Update record corpus.

[BobBorges]

I'll come up with the csv file asap (today-ish).