Formål
?
The aim of the research project Link-Lives is to extend the range of register-research, based on Danish data, from decades to centuries. By combining historical research and data science methods, we reconstruct life-courses and family relations of almost everyone who lived in Denmark from 1787 until the introduction of the modern Danish civil registration system (CPR) in 1968. The project is funded by the Innovation Fund Denmark and the Carlsberg Foundation and runs from 2019 to 2025.
Indhold
?
Link-Lives anvendelsepakke v.2 is the second data release the project launches during the course of the project. It contains 64 million person registrations from three types of sources along with links and life courses created from them. There are transcribed person registrations from: 1) Censuses (the years 1787, 1801, 1834, 1840, 1845, 1850, 1860, 1880, 1885, 1901) originally transcribed by volunteers coordinated by the National Archives (hyperlink to www.ddd.dda.dk). 2) The Copenhagen Burial Register (1861-1911) transcribed by volunteers conducted at Copenhagen City Archives (hyperlink to https://kbharkiv.dk/brug-samlingerne/kilder-paa-nettet/begravelser-i-koebenhavn/begravelser-1861-og-frem/).3) Parish Registers (1813-1917) transcribed in Asia by Ancestry (ancestry link that works).The datasets include core information such as name, age/date of birth, gender, marital status, address, occupation and position in household for most individuals. There is information about family relations and events (births, confirmations, marriages, deaths, arrivals and departures) with dates, places, etc. The data is given in three formats: the original transcription, a version fully standardized and harmonized by Link-Lives and a minimally standardized version to use in our software. Additionally, there are over x million links identifying the same individual person in different datasets. The links are created through three methods:1) A benchmark dataset created by domain experts of 40000 records used for training and testing. 2) Two set of links created through a rule-based algorithm and machine learning implementations of the XG Boost algorithm. 3) Two sets of lifecourses created thee two methods. Given the different origin and ownership of the original datasets, not all the data can be downloaded. Access to the Ancestry data requires permission from Ancestry and the originally transcribed version of Copenhagen City Archives datasets must be obtained from them, where they are freely accessible (see Guide version 2 for details).There will be future releases with more sources, further standardization and different linking methods up to 2025, when the research project Link-Lives ends. By downloading Link-Lives data, you agree to the conditions for use as described in the Link-Lives guide version 2.