Historical Census Record Linkage

Ruggles, Steven; Fitch, Catherine A; Roberts, Evan

doi:10.1146/annurev-soc-073117-041447

Cited by 60 publications

(33 citation statements)

References 83 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…4, 2SH results in more accurate similarities calculated on encoded integer sets compared to both BF and TH. 2SH unlikely results in false matches which will lead to more precise classification of record pairs [7]. We also note that as l increases the similarity difference between the integer and 5.…”

Section: Resultsmentioning

confidence: 88%

“…Preventing false matches and false non-matches are important in many applications of PPRL, such as health or national security, where linking people wrongly or missing links of people that likely are a match can result in difficult to correct biases and thus wrong conclusions [7]. When the length of the bit vectors, l, increases the difference δ between the estimated and the actual Jaccard similarities gets lower.…”

Section: Accuracy Of Similarity Estimationmentioning

confidence: 99%

See 1 more Smart Citation

Secure and Accurate Two-Step Hash Encoding for Privacy-Preserving Record Linkage

Ranbaduge

Christen

Schnell

2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

In order to discover new insights from data, there is a growing need to share information that is distributed across multiple databases that are often held by different organisations. One key task in data integration is the calculation of similarities between records to identify pairs or sets of records that correspond to the same real-world entities. Due to privacy and confidentiality concerns, however, the owners of sensitive databases are often not allowed or willing to exchange or share their data with other organisations to allow such similarity calculations. In this paper we propose a novel privacy-preserving encoding technique that can be used to securely calculate similarities between sensitive values held in different databases. Our technique uses two-step hashing to encode values into an integer set representation that provides strong privacy guarantees and allows accurate similarity calculations. We provide a theoretical analysis of the accuracy and privacy of our encoding technique, and conduct an empirical study on large real databases containing several millions records. Our results show that our technique provides high security against privacy attacks and achieves better similarity accuracy compared to two state-of-the-art encoding techniques.

show abstract

Section: Resultsmentioning

confidence: 88%

Section: Accuracy Of Similarity Estimationmentioning

confidence: 99%

Secure and Accurate Two-Step Hash Encoding for Privacy-Preserving Record Linkage

Ranbaduge

Christen

Schnell

2020

Advances in Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

show abstract

“…There exist many possible mechanisms to perform this kind of record linkage (for a review, see e.g. Ruggles et al, 2018). To narrow the focus down to two of the most promising linking methods, I follow the recommendations by Bailey et al (2017a) and evaluate the mechanism introduced by Ferrie (1996) and the one developed by Abramitzky et al (2012Abramitzky et al ( , 2014 13 So, an individual that is categorized around the age of forty has to be between thirty-five and fortyfive.…”

Section: Samples and Descriptivesmentioning

confidence: 99%

“…Record linkage is widely applied in both historical and contemporaneous research contexts (Ruggles et al, 2018 provide an overview). Consequently, many researchers aim at automating and improving linking procedures 4 or evaluating existing mechanisms (Bailey et al, 2017a;Eriksson, 2017;Massey, 2017).…”

Section: Introductionmentioning

confidence: 99%

Bias in Social Mobility Estimates with Historical Data: Evidence from Swiss Microdata

Favre

2019

SSRN Journal

View full text Add to dashboard Cite

This paper explores a variety of potential issues one has to address when estimating intergenerational mobility with historical data. Many studies are potentially affected by bias originating from individuals emigrating and thus dropping out of the sample, missing information on the life-cycle, and imperfectly linking data sets. Unique panel data on Zurich's citizenry between 1799 and 1926 entail information on true intergenerational links, and allow to follow individuals across the globe and time. This information enables me to explore how father-son mobility estimates are affected by excluding emigrating individuals, occupational patterns over the life-cycle, and linking procedures. The results suggest that focusing on geographically immobile individuals might decrease the estimated level of social mobility. The estimated level of mobility depends on both the father's and the son's age at classification but does not exhibit a monotone trend in the direction of the bias. Most recent linking procedures do not generate significant bias in the sample of Zurich citizens due to the high level of detail of the data combined with a small population size.

show abstract

“…The exponential increase in the resolution of analysis invites scholars to revisit many assumptions about populations of interest, sample weighting, validation or ground-truthing, and measurement. As Ruggles, Fitch, and Roberts (2018) suggest the systematic effort to link repeated observations for social and economic research reaches back to work in the 1930s. But in the last two decades, the focus has shifted onto a larger geographic stage, moving from intensive studies of local and regional settings, to national and international studies of migration, mobility, and population change.…”

Section: Introductionmentioning

confidence: 99%