PERFORMING DATA PROCESSING BASED ON DECISION TREE
Disclosed herein are methods, systems, and apparatus, including computer programs encoded on computer storage media, for data processing. One of the methods includes: determining a set of values in the set of splitting criteria based on the service data, wherein the set of values indicate whether the set of splitting criteria of the burst node are met; encrypting the set of values using a random number, to obtain cyphertext of the set of values; executing a secure data selection algorithm by using the ciphertext of the set of values as input; and executing a secure multi-party computation algorithm by using the random number as input to obtain a prediction result of a decision forest.
1 . (canceled)
2 . A computer-implemented method comprising:
selecting, as a target burst node, a burst node that is associated with service data of a data owner from a decision forest, wherein the decision forest comprises at least one decision tree, and wherein each decision tree comprises at least one burst node and at least two leaf nodes, wherein each burst node includes is associated with a splitting criterion, and wherein each leaf node is associated with a leaf value;
generating a fake splitting criterion for the target burst node;
generating, for the target burst node, a splitting criterion set comprising (i) the fake splitting criterion for the target burst node, and (ii) the splitting criterion that is associated with the target burst node; and
transmitting the splitting criterion set to the data owner.
3 . The method of claim 2 , wherein each burst node in the decision forest corresponds to a data type.
4 . The method of claim 2 , wherein a data type corresponding to the target burst node is the same as the data type corresponding to the service data.
5 . The method of claim 2 , wherein the data owner has all of the service data.
6 . The method of claim 2 , wherein a model owner has part of the service data, and the data owner has another part of the service data.
7 . The method of claim 2 , wherein the decision forest comprises another burst node.
8 . The method of claim 2 , comprising:
saving the splitting criterion that is associated with another burst node and a leaf value corresponding to the leaf node.
9 . A computer-implemented system comprising one or more computers, and one or more computer memory devices interoperably coupled with the one or more computers and having tangible, non-transitory, machine-readable media storing one or more instructions that, when executed by the one or more computers, perform operations comprising:
selecting, as a target burst node, a burst node that is associated with service data of a data owner from a decision forest, wherein the decision forest comprises at least one decision tree, and wherein each decision tree comprises at least one burst node and at least two leaf nodes, wherein each burst node includes is associated with a splitting criterion, and wherein each leaf node is associated with a leaf value;
generating a fake splitting criterion for the target burst node;
generating, for the target burst node, a splitting criterion set comprising (i) the fake splitting criterion for the target burst node, and (ii) the splitting criterion that is associated with the target burst node; and
transmitting the splitting criterion set to the data owner.
10 . The system of claim 9 , wherein each burst node in the decision forest corresponds to a data type.
11 . The system of claim 9 , wherein a data type corresponding to the target burst node is the same as the data type corresponding to the service data.
12 . The system of claim 9 , wherein the data owner has all of the service data.
13 . The system of claim 9 , wherein a model owner has part of the service data, and the data owner has another part of the service data.
14 . The system of claim 9 , wherein the decision forest comprises another burst node.
15 . The system of claim 9 , wherein the operations comprise:
saving the splitting criterion that is associated with another burst node and a leaf value corresponding to the leaf node.
16 . A non-transitory, computer-readable medium storing one or more instructions executable by a computer system to perform operations comprising:
selecting, as a target burst node, a burst node that is associated with service data of a data owner from a decision forest, wherein the decision forest comprises at least one decision tree, and wherein each decision tree comprises at least one burst node and at least two leaf nodes, wherein each burst node includes is associated with a splitting criterion, and wherein each leaf node is associated with a leaf value;
generating a fake splitting criterion for the target burst node;
generating, for the target burst node, a splitting criterion set comprising (i) the fake splitting criterion for the target burst node, and (ii) the splitting criterion that is associated with the target burst node; and
transmitting the splitting criterion set to the data owner.
17 . The medium of claim 16 , wherein each burst node in the decision forest corresponds to a data type.
18 . The medium of claim 16 , wherein a data type corresponding to the target burst node is the same as the data type corresponding to the service data.
19 . The medium of claim 16 , wherein the data owner has all of the service data.
20 . The medium of claim 16 , wherein a model owner has part of the service data, and the data owner has another part of the service data.
21 . The medium of claim 16 , wherein the decision forest comprises another burst node.