Framework for Applying Data Masking and Geo-Perturbation Methods to Household Travel Survey Datasets

摘要

Travel data often require significant resources. Once collected, these are very valuable due totheir extensive and growing utility in transportation policy and travel behavior research. Detailedgeospatial referencing of home, work and other travel destinations are common practice andpermit the integration with other spatially archived data sources, such as land use characteristics,transportation system information, other built environment datasets, and social and economicdata. Public agencies, private consultancies, non-profits and educational institutions may benefitfrom access to original data with applications to areas such as public health, equity,transportation safety and urban planning. But distribution of these important and expensive datais limited by the requirement to protect the confidentiality of survey participants, who areguaranteed anonymity in exchange for participation. Data are often aggregated to a geographiclevel such as Census tracts or transportation analysis zones before disseminating to public, whichlimits the utility of this information. As transportation networks, applications and datasets getmore complex and intelligent, there is a great need to make spatially explicit data available toresearchers. This paper aims to develop an analytical framework to explore systematic ways torelease geographically refined data to a wider range of public users. We evaluate disclosure riskand data utility using variety of assumptions and from the perspective of data intruders tofacilitate the proposed framework.