Lemon Inc.

Other

Shanghai Huali Integrated Circuit Corporation

Beijing Zitiao Network Technology Co., Ltd.

Shanghai Huali Microelectronics Corporation

Beijing Jingdong Shangke Information Technology Co., Ltd.

Beijing Jingdong Century Trading Co., Ltd.

Tiktok Pte. Ltd.

Yan, Qiang

Qiang Yan

Jian Du

Haohao Qian

Bo Jiang

Yongjun Zhao

Jun Tan

Qiuming Huang

Dong Zhang

Sagar Sharma

Yihui Shen

Chaoqi Jiang

Haifeng Niu

Chuyuan Chen

Yongchuan Niu

Li Wang

Xiao Wang

Aihua Li

Yongyue Chen

Jiaqi Hong

Donghang Lu

Dong Yin

Shengli Ge

The present disclosure involves methods, apparatus, and systems for processing group-by operations in databases. In one example, a method includes receiving a query to perform a group-by operation based on first attributes and second attributes of a plurality of entities. The first attributes are private data owned by a first participant of a secure multi-party computation (MPC), and the second attributes are private data owned by a second participant of the secure MPC. The method further includes mapping the first attributes to a first set of numbers, and calculating, based on secure MPC, a secret share of a dataset. The dataset includes the first set of numbers and the second attributes. The method further includes performing the group-by operation including arranging the plurality of entities into groups based on at least the first set of numbers, and aggregating the second attributes of entities in each of the groups.

Processing group-by operations in databases

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating a random noise value for use in providing differential privacy in multiparty computation. One of the method includes receiving, by a first party participating in a multiparty secure computation ('MPC') instance, a share of a first n-bit random binary number and a plurality of function table segments; generating a share of a second n-bit random binary number; calculating a first value based on a comparison between the share of the first and second n-bit random binary numbers; reconstructs an overall value of a comparison between the first and second n-bit random binary numbers; and determining a share of the random noise value based on the overall value and the plurality of function table segments.

Sampling from binomial distribution in secure multi-party computation

Methods and systems for evaluating privacy protection are provided. A method includes determining a first sub-dataset of a first dataset based on a sampling rate, a tolerance, and a first threshold. The method also includes determining a second sub-dataset of the first dataset based on the sampling rate, the tolerance, and a second threshold. The method includes determining a first distance between the first sub-dataset and the first threshold, determining a second distance between the second sub-dataset and the second threshold, generating a first intersection of the first sub-dataset and a second dataset and updating posterior for elements of the first dataset based on the first intersection when the first distance is less than or equal to the second distance, and determining positive membership and negative membership for the elements of the first dataset in the second dataset based on the posterior for the elements of the first dataset.

Statistical measurement with active learning for privacy protection evaluation

Differential privacy composition determination in a secure communication is provided. A method for determining differential privacy composition includes determining a differential privacy configuration including a first privacy parameter and a second privacy parameter, determining a privacy loss distribution, and performing a partial convolution operation of the privacy loss distribution by transforming the privacy loss distribution based on a predetermined parameter, determining an integral range based on the first privacy parameter and the privacy loss distribution, transforming the privacy loss distribution based on the integral range, and performing the partial convolution operation based on the integral range. The method also includes determining the second privacy parameter based on a result of the partial convolution operation, and controlling a dataset based on the differential privacy configuration to limit access to the dataset.

Transformed partial convolution algorithm for composition determination

The present application discloses an epitaxial growth method for an FDSOI hybrid region, comprising: step 1, providing an FDSOI substrate structure, and forming a hard mask layer; step 2, forming a trench in the entire hybrid region, wherein the bottom surface of the trench is below or level with the top surface of the semiconductor body layer; step 3, performing oxidation to form a first oxide layer on the exposed surfaces of the semiconductor body layer and the semiconductor top layer; step 4, fully etching the first oxide layer, and forming an inner sidewall composed of the remaining first oxide layer on the side surface of the trench in a self-aligned manner; and step 5, performing epitaxial growth to form, in the trench, a semiconductor epitaxial layer in contact with the semiconductor body layer.

Epitaxial growth method for FDSOI hybrid region

The present application discloses an epitaxial growth method for an FDSOI hybrid region, comprising: step, providing an FDSOI substrate structure; step, forming a trench; step, performing first isotropic epitaxial growth, wherein a top surface of a first semiconductor epitaxial sublayer is located in a plane between a top surface and a bottom surface of a dielectric buried layer, and a second semiconductor epitaxial sublayer comprise a lateral protruding structure on the side face of the semiconductor top layer; and step, performing second epitaxial growth having a growth rate of the first crystalline face, which is greater than a growth rate of the second crystalline face, wherein the third semiconductor epitaxial sublayer has a chamfered recess near the side face of the semiconductor top layer, finally the lateral protruding structure is located in the chamfered recess.

Methods and systems for secure computation and communication are provided. The method includes transforming identifications of a first dataset using a first transforming scheme, and transforming attributes of the first dataset using a second transforming scheme. The method also includes dispatching the transformed first dataset, receiving a second dataset, transforming identifications of the received second dataset, dispatching the identifications of the transformed received second dataset, and receiving a set of identifications. The method further includes generating a first intersection of the received set of identifications and the transformed received second dataset, generating a first share based on the first intersection, receiving a second share, and constructing a result based on the first share and the second share.

Secure computation and communication

Modelling for efficient resource allocation and/or distribution in secure computation and communication for differential privacy protocols and/or algorithms is provided. A method for allocating a resource of a differentially private system in secure computation and communication includes aggregating attributes from a usage log of the differentially private system for a predetermined period of time, generating a moving aggregation based on the aggregated attributes, training a machine learning model based on the aggregated attributes and the moving aggregation, predicting a distribution of the resource using the trained machine learning model, and allocating the resource based on the predicted distribution. The resource includes a differential privacy parameter.

Resource allocation in secure computation and communication

Dynamic allocation and/or distribution of a resource in secure computation and communication to achieve a desired level of data privacy protection is provided. A method for dynamically allocating a resource of a differentially private system in secure computation and communication includes aggregating attributes from a usage log of the differentially private system for a predetermined period of time, generating a moving aggregation based on the aggregated attributes, optimizing a first attribute of the attributes of the usage log based on the moving aggregation and a set of constraints, generating a distribution of the resource based on the optimized first attribute, and dynamically allocating the resource based on the generated distribution. The resource includes a differential privacy parameter.

Dynamic resource allocator in secure computation and communication

Protecting membership in secure multi-party computation and communication is provided. A method of protecting membership includes generating a padding dataset, up-sampling a first dataset with the padding dataset, transforming and dispatching the first dataset, receiving a second dataset, and performing a private set intersection operation based on the first dataset and the second dataset to generate a third dataset. Each of the first dataset, the padding dataset, and/or the second dataset includes one or more personal identification information for each user or member in the dataset.

Protecting membership in a secure multi-party computation and/or communication

Protecting data privacy in secure multi-party computation and communication is provided. A method of protecting data privacy includes determining a differential privacy configuration, determining a number of iterations based on a first parameter and a second parameter, and for each of the number of iterations generating a random value and a random noise data; generating a first message and a second message; and performing a transfer based on the first message, the second message, and an input data to output one of the first message and the second message. The method also includes generating a first noise data based on the random noise data in each of the number of iterations, generating a first share based on a first dataset and a second dataset, applying the first noise data to the first share, and constructing a result based on the first share and a second share.

Secure multi-party computation and communication

Protecting membership and data in secure multi-party computation and communication is provided. A method of protecting membership and data includes generating a padding dataset. A size of the padding dataset is determined based on a data privacy configuration. The method also includes up-sampling a first dataset with the padding dataset, transforming the first dataset, dispatching the first dataset, performing an intersection operation based on the first dataset and a second dataset to generate a third dataset, generating a first share based on the third dataset, and constructing a result based on the first share and a second share.

Protecting membership and data in a secure multi-party computation and/or communication

Methods and systems for protecting membership privacy for secure computation and communication are provided. The method includes providing a first dataset, determining a number N based on a data privacy configuration, and generating a padding dataset having more than N elements. An intersection of the padding dataset and the first dataset is empty. The method also includes shuffling the padding dataset, up-sampling the first dataset with a first N elements of the shuffled padding dataset, and performing an intersection operation based on the up-sampled first dataset and a received dataset.

Protecting membership for secure computation and communication

Differential privacy composition determination in secure computation and communication of a dataset is provided. A method for differential privacy composition determination includes determining a differential privacy configuration that includes a first privacy parameter and a second privacy parameter, determining a privacy loss distribution, and providing a number of composition operations. The method also includes determining a third privacy parameter and a fourth privacy parameter for a differential privacy composition based on the differential privacy configuration, the privacy loss distribution, and the number of composition operations. The method further includes controlling the dataset based on at least one of the third privacy parameter and the fourth privacy parameter.

Growing community of inventors

Shanghai, China

Qiang Yan