Skip to content

Latest commit

 

History

History
223 lines (150 loc) · 14.1 KB

project_draft.md

File metadata and controls

223 lines (150 loc) · 14.1 KB
Raw

Project draft

Implementation

Step 1: Examination of Materials

The project participants meet IRL at the National Library in Oslo for a one-day workshop (aka hackathon) in September 2018.

Schedule for the kick-off workshop

Time Topic
10.00 – 10.45 Introduction to NorKorr
10.45 – 11.00 Break
11.00 – 13.00 Create an overview of correspondence editions in our collections
13.00 – 14.00 Lunch
14.00 – 15.45 Write a report on what we have collected, assign tasks for further work and decide on further steps
15.45 – 16.00 Wrapping up + coffee

Examination of Materials

We will take a close look at our materials (digital editions of Norwegian correspondences), with regards to:

Materials to depart from

Goal of the workshop is to come up with (1) an overview of digital editions of correspondences that are good candidates for CorrespSearch and (2) identify the work tasks needed to transform (select, format) the metadata to conform to the CorrespSearch CMIF standard (3) assign tasks to the participants and (4) agree on delivery dates for the work packages.

A brief but precise description of (1) will be done, including a list of authors (see Appendix 1), time periods and content of the correspondences to show the potential for future research on this corpus.

CorrespSearch and CMIF

CMIF Standard

CorrespDesc

Step 2: NorKorr to CorrespSearch

The participants prepare a conference paper and a research article about the workflow (material, data formats, distributed hosting, transformations etc.) for NorKorr as an example of preparing large distributed heterogeneous metadata sets for ingestion into CorrespSearch. They explore the potential of such an endeavour and discuss advantages (and challenges) of using decentralized infrastructure and third-party solutions with regards to Norwegian cultural heritage institution’s scope and policies. The workflow is supposed to serve as a basis for further editions of Norwegian correspondences and should stand as an example of re-use of digital data (and especially metadata) provided by large institutions of cultural heritage and research, especially in the Nordic countries.

The paper could be presented at the TEI2019 International Conference in Graz/Austria in September 2019. Alternatively (or additionally), the paper could be given at one of the DH conferences (DHN2019 in early March, DHd2019 in late March, DH2019 in July) or at conferences for edition philology (NNE2019 in September or AG-Edition in February 2020) or Historical Network Research (HNR2019 tba).

The article should be an edited and polished version of the paper and published soon after delivering the paper (aim: 2020, open access journal, international).

Step 3: NorKorr at NB - The Catalogue Data from the Private Archives and Automatic Correspondence Metadata Extraction from Bokhylla.no

This part of the project is tentative. It shall show the potential of the physical and digital collections of the National Library of Norway, especially in regards to exploiting digital humanities methods in metadata and text mining.

Correspondence Material in the Private Archives

Hvem-til-hvem

Tasks:

  • Take a look at the catalogue data and materials and sketch a workflow for making them compatible with the CMIF-Standard used by CorrespSearch
  • Take a look at database behind “Hvem skrev til hvem” and decide if and how to extract data and transform to CMIF-Standard
  • How is correspondence metadata stored today and in the future and how can it be easily fed into CorrespSearch, granted this is wanted by NB

Bokhylla.no

The digitisation efforts of the NB provide an enormous amount of textual data. Included in this are Norwegian scholarly editions of correspondences that formerly were printed and now have been digitized in full text (OCR). In order to complete the NorKorr metadata set, first a selection of scholarly editions of Norwegian correspondences has to be made. Then, from the text of the editions, correspondence metadata has to be (semi-)automatically extracted, stored, and transformed into a CMIF compliant format to be ingested into CorrespSearch.

This is an endeavour of considerable complexity and size and requires expert knowledge in information retrieval from plain text. At this stage, it should be seen as an experiment in retrieving specific textual data from a large collection of text without too much manual interference. The results of this experiment could be presented as a paper or article in the future and the workflow and code could be published for re-use on other full-text collections that contain flat encoded texts (aka plain text).

A list of URNs of digitized scholarly editions of correspondences and letters will be created (by AR) and the raw OCR files will be extracted from bokhylla.no by LJ. The corpus will then be divided into materials that are in the public domain and such that are still under copyright protection. For the letters still under copyright protection, no full-text can be shown outside of the NB, however, amount of words, word lists and topic models can be extracted that should allow for ‘distant reading’ of these letters.

Aims

The aims for the two subprojects in Step 3 are:

  • meaningful re-use of a tool/service developed by NB (HANSKE / Hvem skrev til hvem)
  • accessibility of correspondence metadata that hasn’t been edited yet (or won’t ever be)
  • targeted retrieval of specific data (correspondence metadata from digitized scholarly editions) from bokhylla.no showing the potential of the digital collection even though it is flat encoded
  • large scale show case for “distant reading the network of Norwegian correspondence”

Bibliography

Links

https://www.nb.no/hanske/brev/

https://www.nb.no/hanske/

http://historicalnetworkresearch.org/

https://journal.tei-c.org/index.php/journal/index

https://correspsearch.net/index.xql?l=en

https://correspsearch.net/index.xql?id=participate

http://dh2016.adho.org/abstracts/121

http://correspsearch-test.nodegoat.net/viewer.p/4/136/scenario/1/geo/fullscreen

http://www.tei-c.org/activities/sig/correspondence/

https://github.com/TEI-Correspondence-SIG

http://www.dokpro.uio.no/

http://www.nnedit.org/index.html

http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-correspDesc.html

http://www.tei-c.org/release/doc/tei-p5-doc/en/html/HD.html#HD44CD

Articles

Dumont, Stefan. “CorrespSearch – Connecting Scholarly Editions of Letters.” Journal of the Text Encoding Initiative, no. 10 (December 7, 2016). https://doi.org/10.4000/jtei.1742.

Neuber, Frederike. “CorrespSearch.” Variants. The Journal of the European Society for Textual Scholarship, no. 12–13 (December 31, 2016): 284–85.

Dumont, S. (2016). correspSearch - A Web Service to Connect Diverse Scholarly Editions of Letters. In Digital Humanities 2016: Conference Abstracts. Jagiellonian University & Pedagogical University, Kraków, pp. 175-178.

Appendix 1

Digital Editions of Letters in NB-kilder

Status: 1/8-2018

FRA TIL ÅR PUBLISERT
Collett, Camilla Bjørnson, Bjørnstjerne 1867 NB kilder 2:2/2015
Collett, Camilla Collett, Johan Christian 1849 NB kilder 2:4/2018
Collett, Camilla Collett, Peter Jonas 1841-1851 NB kilder 2:4/2018
Collett, Camilla Didriks, Emilie 1841-1842 NB kilder 2:4/2018
Collett, Camilla Herre, Johanne Caroline 1830 NB kilder 2:4/2018
Collett, Camilla Herre, Marie Emilie 1830 NB kilder 2:4/2018
Collett, Camilla Ibsen, Henrik 1872-1889 NB kilder 2:1/2014
Collett, Camilla Ibsen, Susanna 1872-1894 NB kilder 2:1/2014
Collett, Camilla Lie, Jonas 1863-1884 NB kilder 2:3/2016
Collett, Camilla Petersen, Emma 1839 NB kilder 2:4/2018
Collett, Camilla Wergeland, Amalie Sofie 1839 NB kilder 2:4/2018
Collett, Camilla Wergeland, Laura Augusta 1846 NB kilder 2:4/2018
Collett, Camilla Wergeland, Nicolai 1841-1844 NB kilder 2:4/2018
Collett, Camilla Wergeland, Oscar 1844 NB kilder 2:4/2018

Planned Digital Editions of Letters in NB-kilder

· Correspondence between Kitty Kielland and Arne Garborg, shall be published in 2018

· Letters from Sumatra (Jacob Iversen et al.), shall be published in 2018

· Letters by Just Qvigstad, shall be published in 2019

· Correspondence between Kitty Kielland and Eilif Peterssen, shall be published in 2019(?)

Digital Editions of Letters in NSL/bokselskap.no

Status: 1/8-2018

FRA TIL ÅR PUBLISERT
Bjørnson, Bjørnstjerne Div. danske mottakere (ca 400 brev) 1854-1874 NSL/bokselskap.no 2010
Brandes, Georg Thoresen, Magdalene 1865-1899 NSL/bokselskap.no 2015
Ibsen, Henrik Bjørnson, Bjørnstjerne 1864-1898 HIS/bokselskap.no 2014
Ibsen, Henrik Collett, Camilla 1877-1893 HIS/bokselskap.no 2013
Ibsen, Henrik Lie, Jonas 1879-1900 HIS/bokselskap.no 2012
Kielland, Alexander L. Div. mottakere (ca 1800 brev) 1869-1906 NSL/bokselskap.no 2010
Obstfelder, Sigbjørn Div. mottakere (ca 200 brev) 1884-1900 NSL/bokselskap.no 2016
Skram, Amalie Skram, Erik 1882-1902 NSL/bokselskap.no 2016
Skram, Erik Skram, Amalie 1882-1902 NSL/bokselskap.no 2016
Thoresen, Magdalene Brandes, Georg 1865-1899 NSL/bokselskap.no 2015
Wergeland, Henrik Bekkevold, Amalie 1838-1845 NSL/bokselskap.no 2011

Digital Letters in emunch.no

Status: 11/10-2018

tba

Appendix 2

Zotero Group Library

For scholarly editions of letters and correspondences in the hold of NB. Link to Group Library.

URNs

tba

Selection Criteria

  • has to be letters as part of an actual correspondence (no fictional letters, no religious sermons in the form of letters, no ‘letters to the editor’, no Briefromane)

  • has to be a scholarly edition - although, in the broadest, most inclusive sense

Q: What to do with so called “utvandrerbrev”? There’s a couple of editions with letters from/to Norwegian emigrants (to the US and other places); could be of interest? Perhaps as a subcorpus; I assume that getting authority data for the senders/receivers is difficult in this case.

Q: Should the corpus be limited to “important” people? Pragmatically speaking, limiting the correspondences to “important” people like writers, politicians, cultural and academic elite, nobility could be the easiest way to create a corpus where authority data is available. Letters and correspondences of other people is interesting from a different perspective (local history, cultural history, genealogy etc.) but presumably harder to back with authority data and larger correspondences (aka ‘letters of a lifetime’).