Data labeling method, data labeling device and data labeling equipment

Abstract

The embodiment of the specification discloses a data labeling method, a data labeling device and data labeling equipment. A solution includes: obtaining feature vectors of all to-be-labeled data; using a density clustering algorithm to carry out clustering on all the feature vectors to obtain a plurality of class clusters; carrying out screening in points contained by the class clusters to obtaina core object set and a non-core object set according to density situations of the class clusters; selecting partial core objects in the core object set of the class clusters and partial non-core objects in the non-core object set of the class clusters, and carrying out labeling; and carrying out automatic diffusion labeling on at least partial other points in the class clusters according to a labeling result.

Description

technical field [0001] This description relates to the technical field of computer software, in particular to a data labeling method, device and equipment. Background technique [0002] Data classification and classification is particularly important as the basic capability of big data security work. However, a large number of tables and fields relying on the traditional manual labeling method will bring a lot of workload. In addition, the data security personnel of the general company give priority to investment, and the understanding of the business is relatively limited, so it is difficult to enumerate all data categories, resulting in It is difficult to guarantee the quality of labeling under the large amount of data, which brings great troubles to data classification and classification. [0003] Based on this, a more effective data labeling scheme is needed. Contents of the invention [0004] The embodiments of this specification provide a data labeling method, devi...

AI Technical Summary

Problems solved by technology

However, a large number of tables and fields relying on the traditional manual labeling method will bring a lot of workload. In addition, the data security personnel of the general company give priority to investment, and the understanding of the business is relatively limited, so it is difficult to enumerate all data categories, resulting in It is difficult to guarantee the quality of labeling under the large amount of data, which brings great troubles to data classification and classification

Images