Method and system for adaptive online updating of ad related models

    公开(公告)号:US11544740B2

    公开(公告)日:2023-01-03

    申请号:US15433030

    申请日:2017-02-15

    摘要: The present teaching relates to generating an updated model related to advertisement selection. In one example, a request is obtained for updating a model to be utilized for selecting an advertisement. A plurality of copies of the model is generated. The model is pre-selected based on a performance metric related to advertisement selection. Based on each of the plurality of copies, a candidate model is created by modifying one or more parameters of the copy of the model to create a plurality of candidate models. One of the plurality of candidate models is selected based on the performance metric. The steps of generating, creating, and selecting are repeated until a predetermined condition is met. The model is updated with the latest selected candidate model when the predetermined condition is met.

    Method and system for recommending content items to a user based on tensor factorization

    公开(公告)号:US11315032B2

    公开(公告)日:2022-04-26

    申请号:US15479337

    申请日:2017-04-05

    摘要: The present teaching relates to recommending content items to a user based on tensor factorization. In one example, a request is received for recommending content items to the user. Tensor data related to a plurality of users and a plurality of content items are obtained based on the request. The tensor data is decomposed into a plurality of sub-tensors based on a prior probability distribution. At least one bound is determined for a tensor factorization model that is generated based on the prior probability distribution. One or more items interesting to the user are predicted based on the at least one bound and the plurality of sub-tensors. At least one of the one or more items is recommended to the user as a response to the request.

    Method and system for analyzing a user agent string

    公开(公告)号:US11120004B2

    公开(公告)日:2021-09-14

    申请号:US14411136

    申请日:2014-11-25

    摘要: Method, system, and programs for analyzing user agent string are disclosed. In one example, a user agent string is received. A list of predefined keywords is obtained. Each of the predefined keywords is associated with a type of user agent information. One or more candidate keywords are extracted from the user agent string based on the list of predefined keywords. An extraction pattern is obtained for each of the one or more candidate keywords. A keyword is determined with a version from the one or more candidate keywords based, at least in part, on the extraction patterns for the one or more candidate keywords. The determined keyword with the version represents the type of user agent information in the received use agent string.

    Filtering machine for sponsored content

    公开(公告)号:US11113714B2

    公开(公告)日:2021-09-07

    申请号:US14984853

    申请日:2015-12-30

    IPC分类号: G06Q30/00 G06Q30/02

    摘要: A filtering machine receives sponsored content and filters the sponsored content according to a quality metric generated by quality model circuitry and assigned to the instance of sponsored content. The quality model circuitry generates the quality metric in accordance with historical feedback received about other sponsored content and a collection of quality factors pertaining to the sponsored content. Based on the quality metric for the sponsored content, the filtering machine can effect service of the sponsored content to a user device for display thereon.

    Method and system for distributed deep machine learning

    公开(公告)号:US11087234B2

    公开(公告)日:2021-08-10

    申请号:US15009968

    申请日:2016-01-29

    IPC分类号: G06N20/00

    摘要: The present teaching relates to distributed deep machine learning on a cluster. In one example, a request is received for estimating one or more parameters associated with a machine learning model on a cluster including a plurality of nodes. A set of data is obtained to be used for estimating the one or more parameters. The set of data is divided into a plurality of sub-sets of data, each of which corresponds to one of the plurality of nodes. Each sub-set of data is allocated to a corresponding node for estimating values of the one or more parameters based on the sub-set of data. Estimated values of the one or more parameters obtained based on a corresponding sub-set of data allocated to the node, are received from each of the plurality of nodes. The one or more parameters of the machine learning model are estimated based on the estimated values of the one or more parameters generated by at least some of the plurality of nodes.

    Method and system for rewriting a query

    公开(公告)号:US11086866B2

    公开(公告)日:2021-08-10

    申请号:US15130217

    申请日:2016-04-15

    摘要: The present teaching relates to rewriting a query and providing search results. In one example, a plurality of queries is obtained. For each of the plurality of queries, one or more search results are identified. The one or more search results have been obtained in response to the query and have been previously selected by a user submitting the query. A plurality of titles is obtained. Each of the titles corresponds to one of the one or more search results with respect to one of the plurality of queries. A model is generated based on the plurality of queries and the plurality of titles. The model is to be used for rewriting a query.

    Spark satellite clusters to HADOOP data stores

    公开(公告)号:US10902464B2

    公开(公告)日:2021-01-26

    申请号:US14470704

    申请日:2014-08-27

    IPC分类号: G06Q30/02

    摘要: An advertising and data analysis platform may need to mine through vast amounts of data to come up with insights into advertising effectiveness, and measure and improve the effectiveness of advertising reach. Distributed network data analytics may be applied to ad matching/targeting, such that an in-memory cluster computing environment may be used with advertising data. For example, HADOOP may be utilized for distributed processing of the vast amounts of data and the HADOOP distributed file system (HDFS) is used for organizing communications and storage of that data. Satellite clusters or nodes may be generated that also utilize HDFS. For example, a SPARK or SHARK satellite cluster may be arranged to further utilize the HDFS of the HADOOP clusters.

    METHOD FOR IDENTIFYING MULTIPLE DEVICES BELONGING TO THE SAME GROUP

    公开(公告)号:US20180248844A1

    公开(公告)日:2018-08-30

    申请号:US15856801

    申请日:2017-12-28

    摘要: The technologies described herein identify multiple electronic devices belonging to the same group. A computer system receives, from network accessing applications of a plurality of electronic devices, internet protocol (IP) trajectory information about the network accessing applications via a network. The IP trajectory information includes a user identifier, a list of IP addresses associated with each of the network accessing applications, and timestamps specifying times each of the network accessing applications accesses the network. The computer system identifies and removes commercial IP addresses from the list of IP addresses, analyzes IP trajectory information to retrieve a most commonly used IP address for each of the network accessing applications during a certain period, and determines that different network accessing applications belong to the same group if the most commonly used IP addresses for the different network accessing applications are the same.

    METHODS AND SYSTEMS FOR MONITORING VIEWABLE IMPRESSIONS OF ONLINE CONTENT

    公开(公告)号:US20180218391A1

    公开(公告)日:2018-08-02

    申请号:US15421052

    申请日:2017-01-31

    发明人: Ilya SLAIN

    IPC分类号: G06Q30/02

    摘要: Methods and systems to monitor viewable impressions of online advertisements are provided. A method includes receiving a request to provide an online advertisement for display on a user device. The method further includes providing an online advertisement viewability rules array to an online content publisher system in response to receiving the request. The online advertisement viewability rules array includes a plurality of rules, each rule of the plurality of rules including a definition of when a viewable impression of an online advertisement occurs on a user device. The method further includes receiving, through operation of the online content publisher system, data indicating a viewable impression of the online advertisement has occurred on the user device. The data is generated in response to a determination a viewable impression of the online advertisement has occurred on the user device according to at least one rule of the online advertisement viewability rules array.