TRAINING A SEARCH RESULT RANKER WITH AUTOMATICALLY-GENERATED SAMPLES
    53.
    发明申请
    TRAINING A SEARCH RESULT RANKER WITH AUTOMATICALLY-GENERATED SAMPLES 有权
    用自动生成样本培养搜索结果排名

    公开(公告)号:US20100082510A1

    公开(公告)日:2010-04-01

    申请号:US12243359

    申请日:2008-10-01

    IPC分类号: G06F15/18 G06F7/06 G06F17/30

    CPC分类号: G06N99/005 G06F17/3053

    摘要: A search result ranker may be trained with automatically-generated samples. In an example embodiment, user interests are inferred from user interactions with search results for a particular query so as to determine respective relevance scores associated with respective query-identifier pairs of the search results. Query-identifier-relevance score triplets are formulated from the respective relevance scores associated with the respective query-identifier pairs. The query-identifier-relevance score triplets are submitted as training samples to a search result ranker. The search result ranker is trained as a learning machine with multiple training samples of the query-identifier-relevance score triplets.

    摘要翻译: 搜索结果筛选器可以用自动生成的样本进行训练。 在一个示例性实施例中,用户兴趣从用户与特定查询的搜索结果的交互推断,以便确定与搜索结果的相应查询 - 标识符对相关联的相应关联度得分。 查询标识符 - 相关性分数三元组由与相应查询 - 标识符对相关联的各个相关性得分制定。 查询标识符 - 相关性分数三元组作为训练样本提交给搜索结果筛选器。 搜索结果筛选器被训练为具有查询标识符相关性分数三元组的多个训练样本的学习机器。

    RANKER SELECTION FOR STATISTICAL NATURAL LANGUAGE PROCESSING
    54.
    发明申请
    RANKER SELECTION FOR STATISTICAL NATURAL LANGUAGE PROCESSING 有权
    用于统计自然语言处理的排名选择

    公开(公告)号:US20090125501A1

    公开(公告)日:2009-05-14

    申请号:US11938811

    申请日:2007-11-13

    IPC分类号: G06F7/10

    CPC分类号: G06F17/2715

    摘要: Systems and methods for selecting a ranker for statistical natural language processing are provided. One disclosed system includes a computer program configured to be executed on a computing device, the computer program comprising a data store including reference performance data for a plurality of candidate rankers, the reference performance data being calculated based on a processing of test data by each of the plurality of candidate rankers. The system may further include a ranker selector configured to receive a statistical natural language processing task and a performance target, and determine a selected ranker from the plurality of candidate rankers based on the statistical natural language processing task, the performance target, and the reference performance data.

    摘要翻译: 提供了用于选择用于统计自然语言处理的游戏者的系统和方法。 一种公开的系统包括被配置为在计算设备上执行的计算机程序,该计算机程序包括数据存储器,该数据存储器包括用于多个候选排名者的参考演出数据,该参考演出数据是基于每个测试数据的处理来计算的 多个候选排名。 该系统可以进一步包括配置成接收统计自然语言处理任务和性能目标的排队选择器,并且基于统计自然语言处理任务,性能目标和参考性能来确定来自多个候选排名者的选定队员 数据。

    LIMITED-MEMORY QUASI-NEWTON OPTIMIZATION ALGORITHM FOR L1-REGULARIZED OBJECTIVES
    55.
    发明申请
    LIMITED-MEMORY QUASI-NEWTON OPTIMIZATION ALGORITHM FOR L1-REGULARIZED OBJECTIVES 有权
    用于L1规范化目标的有限存储器QUASI-NEWTON优化算法

    公开(公告)号:US20090106173A1

    公开(公告)日:2009-04-23

    申请号:US11874199

    申请日:2007-10-17

    IPC分类号: G06F15/18

    CPC分类号: G06N99/005

    摘要: An algorithm that employs modified methods developed for optimizing differential functions but which can also handle the special non-differentiabilities that occur with the L1-regularization. The algorithm is a modification of the L-BFGS (limited-memory Broyden-Fletcher-Goldfarb-Shanno) quasi-Newton algorithm, but which can now handle the discontinuity of the gradient using a procedure that chooses a search direction at each iteration and modifies the line search procedure. The algorithm includes an iterative optimization procedure where each iteration approximately minimizes the objective over a constrained region of the space on which the objective is differentiable (in the case of L1-regularization, a given orthant), models the second-order behavior of the objective by considering the loss component alone, using a “line-search” at each iteration that projects search points back onto the chosen orthant, and determines when to stop the line search.

    摘要翻译: 一种使用为优化差分功能而开发的修改方法的算法,但也可以处理L1正则化发生的特殊非差异性。 该算法是L-BFGS(有限存储器Broyden-Fletcher-Goldfarb-Shanno)准牛顿算法的修改,但现在可以使用在每次迭代中选择搜索方向的过程来处理梯度的不连续性,并且修改 线搜索程序。 该算法包括一个迭代优化过程,其中每次迭代大致使目标在目标可微分的空间的约束区域(在L1正则化的情况下,给定的不对称)下的目标最小化,对目标的二阶行为进行建模 通过考虑单独的损失组件,在每次迭代时使用“线搜索”来将搜​​索点投射回所选择的不同,并确定何时停止线搜索。

    Finite-state model for processing web queries
    56.
    发明申请
    Finite-state model for processing web queries 失效
    用于处理Web查询的有限状态模型

    公开(公告)号:US20080183673A1

    公开(公告)日:2008-07-31

    申请号:US11698011

    申请日:2007-01-25

    IPC分类号: G06F17/30

    CPC分类号: G06F17/30864

    摘要: A method of creating an index of web queries is discussed. The method includes receiving a first query representative of one or more symbolic characters and assigning the first query to a first data structure. A first text string representative of the first query is created and assigned to a second data structure. The first and second data structures are stored on a tangible computer readable medium.

    摘要翻译: 讨论了创建Web查询索引的方法。 该方法包括接收表示一个或多个符号字符的第一查询,并将第一查询分配给第一数据结构。 创建表示第一查询的第一文本串并将其分配给第二数据结构。 第一和第二数据结构存储在有形的计算机可读介质上。

    Method and apparatus for distribution-based language model adaptation
    58.
    发明授权
    Method and apparatus for distribution-based language model adaptation 有权
    基于分布式语言模型适应的方法和装置

    公开(公告)号:US07254529B2

    公开(公告)日:2007-08-07

    申请号:US11225543

    申请日:2005-09-13

    IPC分类号: G06F17/27 G06F17/28 G10L15/00

    摘要: A method and apparatus are provided for adapting a language model to a task-specific domain. Under the method and apparatus, the relative frequency of n-grams in a small training set (i.e. task-specific training data set) and the relative frequency of n-grams in a large training set (i.e. out-of-domain training data set) are used to weight a distribution count of n-grams in the large training set. The weighted distributions are then used to form a modified language model by identifying probabilities for n-grams from the weighted distributions.

    摘要翻译: 提供了一种用于使语言模型适应于任务特定领域的方法和装置。 在该方法和装置下,小训练集中的n-gram的相对频率(即任务特定的训练数据集)和大训练集中的n-gram的相对频率(即,域外训练数据集 )用于在大训练集中加权n-g的分布计数。 然后通过从加权分布中识别n克的概率,将加权分布用于形成修改后的语言模型。

    Method and apparatus for compressing asymmetric clustering language models
    59.
    发明授权
    Method and apparatus for compressing asymmetric clustering language models 有权
    用于压缩非对称聚类语言模型的方法和装置

    公开(公告)号:US07231349B2

    公开(公告)日:2007-06-12

    申请号:US10448498

    申请日:2003-05-30

    申请人: Mu Li Jianfeng Gao

    发明人: Mu Li Jianfeng Gao

    IPC分类号: G01L15/06

    摘要: A method and data structure are provided for efficiently storing asymmetric clustering models. The models are stored by storing a first level record for a word identifier and two second level records, one for a word identifier and one for a cluster identifier. An index to the second level word record and an index to the second level cluster record are stored in the first level record. Many of the records in the data structure include both cluster sub-model parameters and word sub-model parameters.

    摘要翻译: 提供了一种方法和数据结构,用于有效地存储非对称聚类模型。 通过存储用于字标识符的第一级记录和两个第二级记录来存储模型,一个用于字标识符,一个用于集群标识符。 第二级记录的索引和第二级集群记录的索引存储在第一级记录中。 数据结构中的许多记录包括集群子模型参数和单词子模型参数。

    Context modeling architecture and framework
    60.
    发明申请
    Context modeling architecture and framework 有权
    上下文建模架构和框架

    公开(公告)号:US20070112546A1

    公开(公告)日:2007-05-17

    申请号:US11253866

    申请日:2005-10-19

    IPC分类号: G06F17/10

    CPC分类号: G06N99/005 G06F9/453

    摘要: A context modeling architecture that includes a context representation portion, which adapted to represent context as features, is provided. The features are specifiable at runtime of an application including the context representation portion.

    摘要翻译: 提供了一种包括上下文表示部分的上下文建模体系结构,其适用于将上下文表示为特征。 这些特征在包括上下文表示部分的应用的运行时是可指定的。