-
公开(公告)号：US20200319927A1
公开(公告)日：2020-10-08
申请号：US16907637
申请日：2020-06-22
Applicant: Alibaba Group Holding Limited
Inventor： Jun Zhou , Xiaolong Li
Abstract: Evaluation results of a plurality of users are received from a plurality of data providers. The evaluation results are obtained by the plurality of data providers evaluating the plurality of users based on evaluation models of the plurality of data providers. A plurality of training samples is constructed by using the evaluation results. Each training sample includes a respective subset of the evaluation results corresponding to a same user of the plurality of users. A label for each training sample is generated based on an actual service execution status of the same user. A model is trained based on the plurality of training samples and the plurality of labels, including setting a plurality of variable coefficients, each variable coefficient specifying a contribution level of a corresponding data provider. Virtual resources to each data provider are allocated based on the plurality of variable coefficients.
-
公开(公告)号：US10789377B2
公开(公告)日：2020-09-29
申请号：US16390147
申请日：2019-04-22
Applicant: Alibaba Group Holding Limited
Inventor： Chaochao Chen , Jun Zhou
Abstract: An item rating and recommendation platform identifies rating data including respective ratings of multiple items with respect to multiple users; identifies user-feature data including user features contributing to the respective ratings of the multiple items with respect to the multiple users; and receives, from a social network platform via a secret sharing scheme without a trusted initializer, manipulated social network data computed based on social network data and a first number of random variables. The social network data indicate social relationships between any two of the number of users. In the secret sharing scheme without the trust initializer, the social network platform shares with the item rating and recommendation platform manipulated social network data without disclosing the social network data. The item rating and recommendation platform updates the user-feature data based on the rating data and the manipulated social network data.
-
Patent Agency Ranking
公开(公告)号：US10789322B2
公开(公告)日：2020-09-29
申请号：US16833021
申请日：2020-03-27
Applicant: Alibaba Group Holding Limited
Inventor： Jun Zhou
IPC: G06F16/955 , G06F16/958 , H04L29/12 , G06F16/00 , G06F16/242
Abstract: A server receives a short link application from a requester. The short link application includes a long link uniform resource locator (URL). The server obtains a database identifier based on the long link URL. The server determines whether a database associated with the database identifier is accessible by the server. In response to a determination that the database associated with the database identifier is accessible by the server, the server obtains a short link URL associated with the long link URL from the database, and transmits the short link URL to the requester.
-
公开(公告)号：US10748090B2
公开(公告)日：2020-08-18
申请号：US16043006
申请日：2018-07-23
Applicant: ALIBABA GROUP HOLDING LIMITED
Inventor： Jun Zhou
IPC: G06Q10/04
Abstract: The present disclosure provides machine-exception handling methods and learning rate adjustment methods and apparatuses. One exemplary method comprises: acquiring a gradient consumption time of a target machine, wherein the gradient consumption time is used to indicate a gradient related time consumed by the target machine in a training process; determining whether the gradient consumption time satisfies a predetermined condition compared with a pre-acquired average consumption time, wherein the average consumption time is used to indicate an average value of the gradient related time consumed by all machines other than the target machine in a cluster in the training process; and determining that the target machine is abnormal if the gradient consumption time satisfies the predetermined condition compared with the average consumption time. The present disclosure addresses the technical problem of high training costs caused by low computation or communication speeds of some machines in a cluster.
-
公开(公告)号：US10691494B2
公开(公告)日：2020-06-23
申请号：US16697913
申请日：2019-11-27
Applicant: Alibaba Group Holding Limited
Inventor： Jun Zhou , Xiaolong Li
Abstract: Evaluation results of a plurality of users are received from a plurality of data providers. The evaluation results are obtained by the plurality of data providers evaluating the plurality of users based on evaluation models of the plurality of data providers. A plurality of training samples is constructed by using the evaluation results. Each training sample includes a respective subset of the evaluation results corresponding to a same user of the plurality of users. A label for each training sample is generated based on an actual service execution status of the same user. A model is trained based on the plurality of training samples and the plurality of labels, including setting a plurality of variable coefficients, each variable coefficient specifying a contribution level of a corresponding data provider. Virtual resources to each data provider are allocated based on the plurality of variable coefficients.
-
公开(公告)号：US20190026816A1
公开(公告)日：2019-01-24
申请号：US16140308
申请日：2018-09-24
Applicant: Alibaba Group Holding Limited
Inventor： Jun Zhou
Abstract: A time-division recommendation method for service objects and an apparatus thereof are provided. The method includes obtaining a user activity log on a service platform; determining time periods of recommendation using the user activity log; separately configuring recommendation strategies for the time periods of recommendation; and recommending service objects to users in the time periods of recommendation using the recommendation strategies correspondingly. The embodiments of the present disclosure are used for satisfying in-depth needs of users, and improving the effects of recommendation of service objects of service platforms.
-
公开(公告)号：US20190026657A1
公开(公告)日：2019-01-24
申请号：US16141886
申请日：2018-09-25
Applicant: Alibaba Group Holding Limited
Inventor： Jun Zhou
Abstract: A distributed cluster training method and an apparatus thereof are provided. The method includes reading a sample set, the sample set including at least one piece of sample data; using the sample data and current weights to substitute into a target model training function for iterative training to obtain a first gradient before receiving a collection instruction, the collection instruction being issued by a scheduling server when a cluster system environment meets a threshold condition; sending the first gradient to an aggregation server if a collection instruction is received, wherein the aggregation server collects each first gradient and calculates second weights; and receiving the second weights sent by the aggregation server to update current weights. The present disclosure reduces an amount of network communications and an impact on switches, and avoids the use of an entire cluster from being affected.
-
-
-
-
-
-
Search Results
37
records
(took 0.017 seconds)
