Improving record linkage with supervised learning for disclosure risk assessment

Abril, Daniel; Navarro‐Arribas, Guillermo; Torra, Vicenç

doi:10.1016/j.inffus.2011.05.001

Cited by 26 publications

(10 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Record linkage is highly used to identify data that being linked, so all datasets under consideration should ideally undergo a matching process prior to the record linkage [13]. Even though there are studies have been carried out in various issues such as software [14,15], data privacy [16,17] and security [13], we suggest it is important to develop a standard set of approach to show the relationship between data, attributes and organizational goals. Even though the term is defined as record linkage but we use the term data linkage as an effort to identify the relationship between data, attributes and organizational goals because we attempt to identify the linkage for each data and attributes that relates to the organizational goals.…”

Section: B Dependency Relationship Of Organizational Data To the Orgmentioning

confidence: 99%

Ontology Based Framework to Configure the Organizational Goal Analysis and Decision-Making

Izhar¹

2017

IJITCS

View full text Add to dashboard Cite

Abstract-Organizational data is essential to assist domain experts and entrepreneurs for decision making process in relation to the organizational goals but the trustworthiness of organizational data in relation to achieving the organizational goals is often questioned because of the vast amount of organizational data available. This paper proposes a methodology to evaluate organizational data that relates to the organizational goals. This refers to the importance of assisting the organization to utilize relevance of organizational data from the vast amount of available data for decision making to the organizational goals. The aim of this paper is to identify dependency relationship of organizational data that match to the organizational goals and to define a metrics as an analysis approach to measure organizational data to be considered relevant to the organizational goals. The experiment is present to implement the propose methodology in the context of Australian economy. The contribution of this paper will serve as a first step in evaluation approach and analysis of organizational data that relates to the achievement of the organizational goals.

show abstract

Section: B Dependency Relationship Of Organizational Data To the Orgmentioning

confidence: 99%

Ontology Based Framework to Configure the Organizational Goal Analysis and Decision-Making

Izhar¹

2017

IJITCS

View full text Add to dashboard Cite

show abstract

“…This research continues the line started by Torra et al [27] where two different optimization problems were presented: one relies on a weighted arithmetic mean and other on the Ordered Weighted Aggregation (OWA) operator. Later on, Abril et al [28] improved the previous weighted mean supervised learning approach and compared it with the current distance-based record linkage technique. They also made an extensive disclosure risk comparison between a large set of well-known protection methods.…”

Section: Related Workmentioning

confidence: 99%

“…In the context of supervised learning approaches for disclosure risk evaluation this was first used in [27,28]. The interest of this definition is that it does not assume that all attributes are equally important in the re-identification process, since there is a weight for each attribute expressing its relevance in the re-identification process.…”

Section: Distance-based Record Linkagementioning

confidence: 99%

“…Section 5.1 reviews the general optimization problem which was first introduced in [28]. Afterwards, in Section 5.2 we present the formalization problems for the weighted mean and the Choquet integral operators, i.e.…”

Section: Supervised Learning For Record Linkagementioning

confidence: 99%

“…The table also includes the number of constraints of the optimization problem in each case. More details and deeper explanations can be found in the following works [28,29]. Table 2: Additional constraints of the weighted mean and Choquet integral.…”

Section: Learning the Optimal Weights Usingmentioning

confidence: 99%

See 2 more Smart Citations

Supervised learning using a symmetric bilinear form for record linkage

2015

Self Cite

View full text Add to dashboard Cite

Record Linkage is used to link records of two different files corresponding to the same individuals. These algorithms are used for database integration. In data privacy, these algorithms are used to evaluate the disclosure risk of a protected data set by linking records that belong to the same individual. The degree of success when linking the original (unprotected data) with the protected data gives an estimation of the disclosure risk.In this paper we propose a new parameterized aggregation operator and a supervised learning method for disclosure risk assessment. The parameterized operator is a symmetric bilinear form and the supervised learning method is formalized as an optimization problem. The target of the optimization problem is to find the values of the aggregation parameters that maximize the number of re-identification (or correct links). We evaluate and compare our proposal with other non-parametrized variations of record linkage, such as those using the Mahalanobis distance and the Euclidean distance (one of the most used approaches for this purpose). Additionally, we also compare it with other previously presented parameterized aggregation operators for record linkage such as the weighted mean and the Choquet integral. From these comparisons we show how the proposed aggregation operator is able to overcome or at least achieve similar results than the other * Corresponding authorEmail addresses: dabril@iiia.csic.es (Daniel Abril ), vtorra@his.se (Vicenç Torra), guillermo.navarro@uab.cat (Guillermo Navarro-Arribas) 1 parameterized operators. We also study which are the necessary optimization problem conditions to consider the described aggregation functions as metric functions.

show abstract

PriTxt: A privacy risk assessment method for text data based on semantic correlation learning

Xiong

Lin

Zhu

et al. 2021

Concurrency and Computation

View full text Add to dashboard Cite

Summary Privacy risk assessment plays a fundamental role in privacy preservation, as it determines the extent to which subsequent processing (such as generalization and obfuscation), should be applied to the sensitive data. However, most existing works on privacy risk assessment have focused on structured data, while unstructured text data remain relatively underexplored due to the complexity of natural language. In this article, we propose a novel method, PriTxt, for evaluating the privacy risk associated with text data by exploiting the semantic correlation. Using definitions derived from the General Data Protection Regulation (GDPR), a de facto standard of privacy preservation in practice, PriTxt first defines the private features that related to individual privacy in order to locate the sensitive words. By using the word2vec algorithm, a word‐embedding model is further constructed to identify the quasi‐sensitive words that are semantically correlated to the private features. The privacy risk of a given text is finally evaluated by aggregating the weighted risks of the sensitive and the quasi‐sensitive words in the text. Experiments on real‐world datasets demonstrate that the proposed PriTxt is effective for conducting risk assessment on text data and further outperforms the traditional methods.

show abstract

Improving record linkage with supervised learning for disclosure risk assessment

Cited by 26 publications

References 21 publications

Ontology Based Framework to Configure the Organizational Goal Analysis and Decision-Making

Ontology Based Framework to Configure the Organizational Goal Analysis and Decision-Making

Supervised learning using a symmetric bilinear form for record linkage

PriTxt: A privacy risk assessment method for text data based on semantic correlation learning

Contact Info

Product

Resources

About