Linking Dutch Civil Certi cates

Abstract

Finding and linking di erent appearances of the same entity in an open Web setting is one of the primary challenges of the Semantic Web. In social and economic history, record linkage has dealt with this problem for a long time, linking historical individual records at a local database level. With the advent of semantic technologies, Knowledge Graphs containing these records have been published, raising the need for large-scale linking techniques that consider the particularities of historical individual linking. In this paper we focus on our current investigation of such techniques to link the Dutch civil certi cates in the LINKS/CLARIAH project. We describe the production of the LINKS Knowledge Graph, and we show its potential at answering domain research questions through its large number of owl:sameAs links. 5

Keywords

• linked data
• digital humanities
• civil certi cates linking

-

Finding and linking equivalent entities (persons, places, events, concepts) on the Web is one of the most important challenges of a Semantic Web of Linked Data. The distributed data publishing paradigm and the scale of the Web exacerbate this problem; various approaches have been proposed to address it, including heuristic-based linking (e.g. string similarity) [ 12 ], cluster-similarity linking [ 11 ], and deep learning-based knowledge graph completion [ 14 ]. The goal is to produce identity links that use the owl:sameAs or skos:exactMatch predicates so data consumers are aware of identity clusters and classes [ 3 ].

Interestingly, the problem has been dealt with in other elds of research; in particular in economic and social history. There, record linkage is a challenging and active area of research, as shown in a recent Historical Methods special issue on the subject [ 19 ]; and is becoming ever more important in economic and social history. Mass digitisation of archival material means that further insight can be obtained by linking individuals and households across di erent records, especially now that sources with complete population coverage are becoming 5 Copyright ©2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). available. Historical civil certi cates are the authoritative sources of birth, marriage and death events in municipality registers, and allow for the reconstruction of lives of the past [ 4 ][ 23 ]. In the Netherlands, the LINKS project [ 15 ] has shown that this reconstruction is, however, often very challenging, as there is generally no ground truth. Individuals are not actively followed over time, but observed during the registration of a vital event. As a result, it is unclear whether, where, and when an individual can be observed. It is not even certain whether follow-up is available at all, because individuals could migrate out of the region of observation [ 4 ]. To complicate matters further, large quantities of historical certi cates have been indexed, which gives rise to data entry errors. These spelling mistakes can be hard to deal with, as twins and other multiple births often receive similar names. Furthermore, rst names were often reused in families to \replace" earlier-born, deceased siblings. Finally, civil servants were known to indicate non-standard mutations, such as name changes, acknowledgement of children, and divorces as side notes. As a result, very important relational information is often not standardised [ 23 ].

In this paper, we summarise our e orts in the LINKS and CLARIAH projects to overcome these challenges, and link the appearance of the same person in 1.5 million birth (1812{1919), marriage (1812{1944) and death (1812{1969) certi cates in the Dutch province of Zeeland. Speci cally, our contributions are: { A description of the LINKS knowledge graph production process by using standard semantic technologies (Section 4) { A highly scalable certi cate linking method based on e cient string similarity through Levenshtein automaton (Section 5) { A preliminary evaluation based on SPARQL queries that use such links (Section 6)

In the next sections, we survey related work (Section 2), describe the original dataset (Section 3), explain our contributions (Sections 4, 5 and 6), and conclude (Section 7). 2

Related Work

Historical record linkage generally requires a previous e ort on digitising large amounts of individual-level historical records, a goal shared by projects like NAPP/IPUMS [ 9 ], the Balsac Population Database6, the Utah Population Database7, Familysearch8, the Scottish Longitudinal Study9, Digitising Scotland10, the Nor6 http://balsac.uqac.ca/english/ 7 https://uofuhealth.utah.edu/huntsman/utah-population-database/ 8 https://www.familysearch.org 9 https://sls.lscs.ac.uk/ 10 https://digitisingscotland.ac.uk/ way Historical Population Register11, Link-Lives12, POPLINK/DDB13, the Scanian Economic Demographic Database (SEDD)14, the North Orkney Population History Project[ 13 ] and Death and Burial Data in Ireland 1864-192215. Linking individuals in the US 1850 and 1860 census is generally considered one of the earliest e orts [ 8 ], and similar approaches for Canada [ 2 ] and Sweden [ 25 ] have followed. [ 17 ] provides a critical review of these and other historical record linkage e orts, with a focus on US data. We share with these e orts a focus on string-based comparison linkage. In other projects (e.g. Digitising Scotland) the goal is to perform group-level linkage as well [ 1 ]. Recent machine learning approaches have gotten a lot of traction in the eld [ 6 ]. For example, recent work on historical US census data uses manually labelled data from familysearch.com as training data [ 18 ]. In the Netherlands, earlier work on Dutch civil certi cates focuses on methodological aspects of record linkage [ 21 ]. The work done in the LINKS project [ 15 ] constitutes a basis for our contribution. 3

Dataset

The digitised civil registry consists at the moment of 27.5 million certi cates. In total, there are 10.3 million birth certi cates, 4.4 marriage certi cates, and 12.7 million death certi cates in the digitised registry at the International Institute of Social History16. The number of available birth and death certi cates di ers strongly as, due to privacy laws, only death certi cates that are more than 50 years old are available for research. Birth certi cates become available with a 100-year delay, marriage certi cates with a 75-year delay, and death certi cates with a 50-year delay. For the moment, the experiments in this paper are restricted to the civil registries produced in the Zeeland region. This dataset of Zeeland civil registries, known as LINKS Zeeland cleaned 2016 01 [ 16 ], consists of 1.5 million certi cates, which represents 5.5% of the total certi cates. Speci cally, there are 698,285 birth certi cates (6.7% of the total birth certi cates), 193,921 marriage certi cates (4.4%), and 665,999 death certi cates (5.2%). This dataset is cleaned, standardised and distributed in a restricted manner [ 15 ] in the form of three CSV les: 1. Locations: containing the locations that show up in the civil certi cates, describing the municipality, province, region and the country of a location.

This le consists of 6 columns and 2,456 rows. 2. Registrations: containing general data from a certi cate registration which exceed the individual level, such as the date and place of birth, marriage or 11 https://www.rhd.uit.no/nhdc/hpr.html 12 https://link-lives.dk/ 13 https://www.umu.se/en/centre-for-demographic-and-ageing-research/ databases/parish-registers-databases/ 14 https://www.ed.lu.se/databases/sedd 15 https://www.dbdirl.com/ 16 https://iisg.amsterdam/en death. This le consists of 10 columns and 1,558,205 rows, with each row representing a single registration in the Zeeland province. 3. Persons: containing all appearances of persons. In general every birth certi cate generates records for three persons (newborn child, mother and father), a marriage certi cate generates minimally six person records (bride with her parents and the groom with his parents) and a death certi cate generates three or four person records (deceased, father, mother and possibly a spouse). This le consists of 33 columns and 5,526,393 rows. 4

LINKS Knowledge Graph

The process of converting the CSV les of the LINKS dataset into a Knowledge Graph consists of three steps. Firstly, we manually design a model for describing and enriching the civil registries data, following Linked Data best practices. Secondly, we transpose the CSV data into an RDF Knowledge Graph, according to our designed model. Finally, we make the graph available for browsing and querying in an e cient manner. 4.1

Designing the civil registries schema

For modelling the civil registries data, we designed a new simple model that reuses, whenever possible, existing vocabularies. This model is presented in Figure 1, and has four main components: - Civil Registrations. The rst component (concepts coloured in brown) describes each civil registration (birth, marriage, or death certi cate), listing its identi er, its sequential number, the location, and date of the registration. - Life Events. The second component (in green) describes the actual life events (birth, marriage, or death event), listing the main individuals involved in this event, the location and the date of this event. In this model, a distinction is made between the civil registration and their associated life events, as certain civil registrations can be produced in di erent dates and locations from where the life event actually happened. - Individuals. The third component (in blue) describes each individual involved in these life events, listing their names, sex, civil status, and birth dates. - Locations. The nal component (in orange) describes the location where each life event has happened and the location where it was registered. In this component, information regarding the municipality, the province, the region, and the country can be available. 4.2

Transposing the data to RDF

For converting the Zeeland dataset to a Knowledge Graph, we use the tool CoW (CSV on the Web converter)17. This batch tool, developed within the CLARIAH 17 https://csvw-converter.readthedocs.io/ project [ 10 ], allows the conversion of datasets expressed in CSV. It uses a JSON schema expressed using an extended version of the CSVW standard, to convert CSV les to RDF in scalable fashion. In the case of the Zeeland dataset, we run the conversion process separately for the three CSV les, by manually designing the JSON schema for Locations18, Registrations19, and Persons20. This manual design of these JSON les allows us to transpose the data according to the model presented in Figure 1. After designing a JSON le for each of the three CSV les, we use the command line to convert each of these les separately. For instance, having both the Locations.csv le with its associated JSON schema Locations.csv-metadata.json in the same directory, the following command is su cient to convert the data to RDF, creating the RDF le Locations.nq encoded in the N-Quads format. $ c o w t o o l c o n v e r t L o c a t i o n s . c s v

The conversion process takes 30 seconds for the le Locations, 100 minutes for Registrations, and around 5 hours for Persons on a SSD disk, with 64GB of memory. 18 https://raw.githubusercontent.com/CLARIAH/wp4-civreg/master/json/ locations.csv-metadata.json 19 https://github.com/CLARIAH/wp4-civreg/blob/master/json/registrations.

csv-metadata.json 20 https://github.com/CLARIAH/wp4-civreg/blob/master/json/persons.

csv-metadata.json

Accessing the RDF knowledge graph

Combining the three resulted N-Quads les results in the LINKS knowledge graph, composed of 58,513,388 triples. This knowledge graph can be accessed online through Druid21, the CLARIAH instance of the TriplyDB triple store22. Druid allows the storage of knowledge graphs, and provides tools to browse, query and visualise our data. For privacy reasons, the LINKS knowledge graph is uploaded as a private dataset on Druid, restricting its access23 to members of the LINKS organisation24 on Druid. We provide publicly accessible links to resources of the knowledge graph when possible.

In addition to accessing the LINKS knowledge graph through the Druid Web hub, authorised users of the LINKS knowledge graph can also access this dataset locally. For enabling easy and e cient access on a normal local machine, we convert the LINKS knowledge graph from N-Quads to HDT (Header, Dictionary, Triples) [ 7 ]. This compact data structure and binary serialisation format for RDF keeps big datasets compressed to save space while maintaining search and browse operations without prior decompression. Converting the LINKS knowledge graph into HDT consists of two simple steps: (i) merge the three RDF N-Quads les into one larger N-Quads le, (ii) convert the resulting merged le to HDT using the rdfhdt library25. 5

Certi cate Linkage

For linking Dutch civil registries, we heavily rely on the string similarity between individuals' names. This is motivated by the high quality of the registered names in most civil certi cates, and the limited spelling variation between di erent civil certi cates for the same individual. An example of such quality maintenance can be observed in marriage registrations, where both the bride and the groom are required to bring their own birth certi cates when registering their marriage. Moreover, married women in the Netherlands keep their own family name in the civil certi cates, which highly facilitates the problem at hand. In the case of death registrations, they are generally registered by next of kin |parents, spouses, children, or siblings| which also highly limits variations in name spelling [ 23 ].

Similarity between two names can be measured in several ways, such as calculating the Levenshtein, Jaccard, or Jaro-Winkler distances. In this work, we take the Levenshtein distance as a basis for matching individuals in civil certi cates. This distance measures the number of single character edits (insertions, deletions or substitutions) required to change one name into the other. The standard algorithm for calculating the Levenshtein distance between two names was proposed by Wagner and Fisher [ 24 ], but can lead to a quadratic time complexity. 21 https://druid.datalegend.net/ 22 https://triply.cc/ 23 https://druid.datalegend.net/LINKS/links-zeeland/ 24 https://druid.datalegend.net/LINKS 25 http://www.rdfhdt.org/manual-of-the-java-hdt-library/ In this work, as we aim to match individuals from a list of millions of certi cates to individuals in another large list of certi cates, the standard approach (or its variants) of calculating the Levenshtein distance by comparing each pair of certi cates is not feasible, as the time complexity of the approach can grow exponentially with the size of the given lists. Therefore, we adopt the approach and the library proposed by Dylon Devo26, based largely on the work of Schulz and Mihov [ 22 ], for the fast selection of candidate individuals within a certain Levenshtein distance. In this approach, the list of target individuals are indexed as a Minimal Acyclic Finite-State Automata (MA-FSA), where a Levenshtein transducer is initialised according to a maximum distance speci ed by the user. When a name is given as a source query with a maximum accepted Levenshtein distance, the states of the Levenshtein automaton corresponding to that name are constructed on-demand as the automaton is evaluated. According to its author, this approach allows to nd for a given name n all candidate names in a list M in linear time on the length of n, and not on the size M . In the following, we describe how we deploy this approach for matching newborns registered in birth certi cates to their marriage certi cates. The general process remains unchanged for other types of linkage, where only the roles of the considered individuals and the link's timeline consistency are adapted accordingly. 5.1

Approach

Finding the marriage certi cate of a certain newborn, when applicable, requires matching three individuals: (i) the newborn in the birth certi cate with the bride or groom of a certain marriage certi cate, (ii) the newborn's mother with the bride's or groom's mother, (iii) the newborn's father with the bride's or groom's father. Once a match, according to a maximum Levenshtein distance, between the three individuals of a birth certi cate and a marriage certi cate is found, we check whether the logical timeline is respected. Only when a match between two certi cates based on the three individuals is found, with a correct logical timeline, a match between the three individuals is registered in the Knowledge Graph. Speci cally, our approach for matching newborns to a marriage certi cate can be divided into 5 main steps: 1. Create six indices, with each index representing a MA-FSA containing the list of all full names of a certain role in marriage certi cates. For instance, the index of the role "bride" contains the full names ( rst name + last name) of all women individuals that got married (i.e. role of bride). For each of these indices, a Levenshtein transducer is initialised according to a maximum Levenshtein distance, given by the user. 2. Create six Key-Value databases, with each database covering a single role r in the marriage certi cate. A key in a database represents a full name f n, and the value represents a list of marriage certi cate identi ers that have for the role r an individual with the name f n. For instance, the entry \Anna 26 https://github.com/universal-automata/liblevenshtein-java/

Aartsen" ! f123323,232344g indicates that both these certi cates have a bride registered with the full name \Anna Aartsen". While such information can be directly queried from the Knowledge Graph, Key-Value databases are a better mean for frequent read requests. In particular, we rely on the RocksDB27 disk-based Key-Value database. 3. Find marriage certi cate candidate(s) for each birth certi cate. For this, we rstly search for the full name of the newborn in the index of the bride or the groom. Considering that the newborn is a girl, this step retrieves a list of candidate names Cnewborn from the bride index, representing a spelling variation within the maximum Levenshtein distance speci ed by the user. If Cnewborn is not empty, we retrieve from the bride's Key-Value database the list of candidate certi cates Enewborn that contain this candidate's name. In the case where Cnewborn contains several candidates, the result will be the union of all returned Enewborn for each candidate. The same process is applied when searching for the full name of the newborn's mother and father in the bride's mother and father indices, respectively returning a list of candidate certi cates Emother and Efather. 4. Filter resulting candidates. Since in the majority of cases, a newborn is expected to have the same registered parents during marriage, we require the match between the birth and the marriage certi cates to be based on the three individuals. Therefore, the preliminary marriage candidates consists of the intersection of Enewborn, Emother and Efather. Finally, out of these preliminary candidates only those that respect the logical timeline are considered. In this case consisting of matching a newborn to a bride or a groom, we expect that the marriage certi cate is registered at least 14 years, and at most 70 years, after its matched birth registration. 5. Save links, in two formats for respecting the preferences of most researchers: (a) CSV le consisting of the birth certi cate identi er, the matched marriage certi cate identi er, with the link metadata consisting mainly of the Levenshtein distance between each matched individual in these certi cates, and time di erence between both registrations, (b) N-Quads le consisting of owl:sameAs links between each matched individual, with each link being asserted in a di erent named graph for describing its context. For instance, the statement h iisg:newbornURI, owl:sameAs, iisg:brideURI, iisg:graph/birthToMarriage/0-2-1 i indicates that the identity link between these two individuals was detected based on a Levenshtein distance of 0 between the newborn and bride's name, a Levenshtein of 2 between their mothers' names, and 1 for the fathers' names. 5.2

Experiments

For testing the scalability of our approach, we evaluated our matching approach on the Zeeland dataset described in Section 3. We rstly evaluated the process 27 http://rocksdb.org

NewbornToPartner PartnerParentsToCouple pLeMrevIaenxndismihvutidemiunal NuLminbkesr of (Rinunmtiimnse) NuLminbkesr of (Rinunmtiimnse) 1 271,230 5 205,477 2 2 289,937 18 224,785 8 3 310,232 74 244,343 25 Table 1: Results of matching newborns in marriage certi cates to brides/grooms (newbornToPartner), and matching parents of brides/grooms in marriage certi cates to their own marriage certi cate (partnerParentsToCouple). of matching newborns in marriage certi cates to brides/grooms in marriage certi cates, and then evaluated the process of matching parents of brides/grooms in marriage certi cates to their own marriage certi cate. Table 1 shows that matching civil registries of a Dutch province takes no more than a few minutes28, with the runtime increasing as the maximum Levenshtein distance per individual increases. It also shows that even with a maximum Levenshtein distance of 1, there is a signi cant overlinking, since the number of detected links (271,230) is larger than the number of marriage certi cates in this dataset (193,921). Therefore, indicating that a number of marriage certi cates were matched to multiple birth certi cates. The source code of this approach is publicly available29. 6

Preliminary Evaluation (Use Cases)

While we expect that a dataset containing information on every Dutch person born in the period 1812{1919 and their family relations will be useful to many 28 Experiments conducted on MacBook Pro, with SSD disk and 16GB of memory 29 https://github.com/CLARIAH/wp4-links

1880 1890 1900 1910 year researchers, it will be especially valuable to demographic, social, and economic historians working with individual-level data. One issue in particular that it can address, is bias in results due to migration.

The key issue there is that, currently, many analyses are based on records from one locality (a village, town, or province). In other words, out-migrants are left out of the data. This is a problem because migrants are di erent from the rest of the population. For example, according to Ruggles [ 20 ] they had di erent ages at marriage and life expectancies. A comparison of the civil registry of Zeeland and a smaller population register data set (HSN) that follows individuals as they move, has also shown that the di erences between such datasets can be explained by the exclusion of migrants out of Zeeland [ 5 ].

The new data created here can take substantial steps to resolve this issue. Only international out-migrants can now go missing, which is a far smaller share of the data. The query30 in Listing 1.1 shows how migrants and non-migrants are easily identi ed in the data. To do this, we compare the location of a marriage with that of both the bride's and the groom's parents' marriage. The results of this query ( gure 2) show that the share of non-migrants between 1840 and 1910 falls from 80 to 72 percent, which means that by the start of the twentieth century, nearly a fourth of the couples moved between their marriage and that of their child. Linking a civil registry for the entire Netherlands as done here allows us to include this large group in future analyses. 30 https://github.com/CLARIAH/wp4-queries-links/blob/master/ marriage-locations.rq

Conclusion

In this work, we described our production process of the LINKS Knowledge Graph, containing civil certi cates of the Dutch province of Zeeland. We presented our approach for linking all these certi cates, within 5 minutes on a regular laptop, and showed how such links can be exploited for conducting demographic analyses using SPARQL. This work is in the process of being extended to cover all certi cates of the Netherlands, enabling larger and more valuable demographic, social and economic analyses.

References

1. Akgun, , Dearle , A. , Kirby , G. , Garrett , E. , Dalton , T. , Christen , P. , Dibben , C. , Williamson , L. : Linking Scottish vital event records using family groups . Historical Methods: A Journal of Quantitative and Interdisciplinary History 0 ( 0 ), 1 { 17 (Mar 2019 ), https://doi.org/10.1080/01615440. 2019 .1571466

2. Antonie , L. , Inwood , K. , Lizotte , D.J. , Andrew Ross , J.: Tracking people over time in 19th century Canada for longitudinal analysis . Machine Learning 95 ( 1 ), 129 {146 (Apr 2014 ). https://doi.org/10.1007/s10994-013-5421-0, https://link.springer. com/article/10.1007/s10994-013-5421-0

3. Beek , W. , Raad , J. , Wielemaker , J., van Harmelen , F. : sameas. cc: The closure of 500m owl: sameas statements . In: European semantic web conference . pp. 65 { 80 . Springer ( 2018 )

4. Van den Berg, N., Van Dijk , I.K. , Mourits , R.J. , Slagboom , P.E. , Janssens , A.A.P.O. , Mandemakers , K. : Families in comparison: An individual-level comparison of life-course and family reconstructions between population and vital event registers . Population Studies: A Journal of Demography pp. 1 { 20 ( 2020 )

5. Berg , N.v.d., Dijk , I.K.v. , Mourits , R.J. , Slagboom , P.E. , Janssens , A.A.P.O. , Mandemakers , K. : Families in comparison: An individual-level comparison of life-course and family reconstructions between population and vital event registers . Population Studies 0 ( 0 ), 1 { 20 (Feb 2020 ), https://doi.org/10.1080/00324728. 2020 . 1718186

6. Feigenbaum , J.J. : Multiple Measures of Historical Intergenerational Mobility: Iowa 1915 to 1940 . The Economic Journal 128 ( 612 ), F446{F481 (Jul 2018 ), https: //onlinelibrary.wiley.com/doi/abs/10.1111/ecoj.12525

7. Fernandez , J.D. , Mart nez-Prieto, M.A. , Gutierrez , C. , Polleres , A. , Arias , M. : Binary rdf representation for publication and exchange (hdt) . Journal of Web Semantics 19 , 22 { 41 ( 2013 )

8. Ferrie , J.P.: A New Sample of Males Linked from the Public Use Microdata Sample of the 1850 U.S. Federal Census of Population to the 1860 U.S. Federal Census Manuscript Schedules . Historical Methods: A Journal of Quantitative and Interdisciplinary History 29 ( 4 ), 141 {156 (Oct 1996 ), https://doi.org/10.1080/ 01615440. 1996 .10112735

9. Goeken , R. , Huynh , L. , Lynch , T.A. , Vick , R.: New Methods of Census Record Linking . Historical Methods: A Journal of Quantitative and Interdisciplinary History 44 ( 1 ), 7 { 14 (Jan 2011 ), https://doi.org/10.1080/01615440. 2010 .517152

10. Hoekstra , R. , Meron~o-Pen~uela, A. , Rijpma , A. , Zijdeman , R. , Ashkpour , A. , Dentler , K. , Zandhuis , I. , Rietveld , L. : The datalegend ecosystem for historical statistics . Journal of Web Semantics 50 , 49 { 61 ( 2018 )

11. Idrissou , A.K. , Hoekstra , R., Van Harmelen , F. , Khalili , A. , Van den Besselaar, P.: Is my: sameas the same as your: sameas? lenticular lenses for context-speci c identity . In: Proceedings of the Knowledge Capture Conference . pp. 1 { 8 ( 2017 )

12. Isele , R. , Jentzsch , A. , Bizer , C. : Silk server-adding missing links while consuming linked data . In: Proceedings of the First International Conference on Consuming Linked Data-Volume 665 . pp. 85 { 96 . CEUR-WS. org ( 2010 )

13. Jenning , J.A. , Sparks , C.A. , Murtha , T. : Interdisciplinary approach to spatiotemporal population dynamics:the north orkney population history project . Historical Life Course Studies pp. 27 { 51 ( 2019 ), http://hdl.handle.net/10622/ 23526343-2019-0002?locatt=view:master

14. Lisena , P. , Meron~ o-Pen~uela, A. , Troncy , R.: MIDI2vec: Learning MIDI Embeddings for Reliable Prediction of Symbolic Music Metadata . Transactions of the International Society for Music Information Retrieval ( 2020 ), under review

15. Mandemakers , K. , Laan , F. : LINKS dataset Genes Germs and Resources , WieWasWie Zeeland, Civil Certi cates, version 2017 .01. International Institute of Social History, Amsterdam

16. Mandemakers , K. , Laan , F. : LINKS-Zeeland challenge , WieWasWie Zeeland , Civil Certi cates, version 2016 . International Institute of Social History, Amsterdam

17. Massey , C.G. : Playing with matches: An assessment of accuracy in linked historical data . Historical Methods: A Journal of Quantitative and Interdisciplinary History 0 ( 0 ), 1 { 15 (Mar 2017 ), http://dx.doi.org/10.1080/01615440. 2017 .1288598

18. Price , J. , Buckles , K. , Van Leeuwen , J. , Riley , I. : Combining Family History and Machine Learning to Link Historical Records . Tech. Rep. w26227, National Bureau of Economic Research , Cambridge, MA (Sep 2019 ). https://doi.org/10.3386/w26227, http://www.nber.org/papers/w26227.pdf

19. Rijpma , A. , Cilliers , J. , Fourie , J.: Record linkage in the Cape of Good Hope Panel . Historical Methods: A Journal of Quantitative and Interdisciplinary History 0 ( 0 ), 1 { 16 (Feb 2019 ). https://doi.org/10.1080/01615440. 2018 . 1517030 , https://doi. org/10.1080/01615440. 2018 .1517030

20. Ruggles , S. : Migration, Marriage, and Mortality: Correcting Sources of Bias in English Family Reconstitutions . Population Studies 46 ( 3 ), 507 {522 (Nov 1992 ), https://doi.org/10.1080/0032472031000146486

21. Schraagen , M.P. , others: Aspects of record linkage . Ph.D. thesis , Leiden Institute of Advanced Computer Science (LIACS), Faculty of Science , Leiden University ( 2014 ), https://openaccess.leidenuniv.nl/handle/1887/29716

22. Schulz , K.U. , Mihov , S. : Fast string correction with levenshtein automata . International Journal on Document Analysis and Recognition 5 ( 1 ), 67 { 85 ( 2002 )

23. Vulsma , R.F. : Burgerlijke stand en bevolkingsregister. Centraal Bureau voor Genealogie, 's-Gravenhage

24. Wagner , R.A. , Fischer , M.J.: The string-to-string correction problem . Journal of the ACM (JACM) 21(1) , 168 { 173 ( 1974 )

25. Wisselgren , M.J. , Edvinsson , S. , Berggren , M. , Larsson , M. : Testing Methods of Record Linkage on Swedish Censuses. Historical Methods: A Journal of Quantitative and Interdisciplinary History 47 ( 3 ), 138 {151 (Jul 2014 ), https://doi.org/ 10.1080/01615440. 2014 .913967