Paypal, Inc.

Fuyuan Kang

Jie Huang

Gaoyuan Wang

Zelin Yan

Techniques are discussed for preparing and executing scanning plans for particular types of information, including personally identifiable information. A user indicates one or more datastores to be scanned for the particular type of information. A scanner determines scan objectives for the scanning plan and classifiers for use in scans conducted according to the scanning plan. The scanner estimates scan performance metrics and scan quality metrics. The scanner presents estimated results for the scanning plan based on the selected classifiers, scan objectives, estimate scan performance metrics, and estimated scan quality metrics. The user can modify the set of scan objectives or select between alternative sets of scan objectives. The scanning plan may be performed iteratively and the results of previous scan may be used to adjust classifiers or scan objectives to be used in subsequent scans.

Scanning for information according to scan objectives

Methods and systems are presented for performing data scanning in a heterogenous data storage environment. A data scanning process on multiple different data storages may be performed based on a data scanning request, which specifies certain data of interest to be detected during the scanning. First, sampled data records are scanned from each of the data storages. Based on analyzing the data in the sampled data records and the metadata associated with each data storage, a fingerprint may be generated for each of the data storages that indicates a likelihood that the data storage includes the data of interest. A second scan may be performed on the data storage only if the fingerprint exceeds a threshold. The data of interest may be detected in the data storages during the second scan of the selected data storages.

Growing community of inventors

Shanghai, China

Fuyuan Kang