公开(公告)号：US07933847B2

公开(公告)日：2011-04-26

申请号：US11874199

申请日：2007-10-17

申请人： Galen Andrew ,  Jianfeng Gao

发明人： Galen Andrew ,  Jianfeng Gao

IPC分类号： G06F15/18 ,  G06F17/27 ,  G06N3/08 ,  G10L15/14 ,  G10L15/00 ,  G10L15/18

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。

公开(公告)号：US20090125501A1

公开(公告)日：2009-05-14

申请号：US11938811

申请日：2007-11-13

申请人： Jianfeng Gao ,  Galen Andrew ,  Mark Johnson ,  Kristina Toutanova

发明人： Jianfeng Gao ,  Galen Andrew ,  Mark Johnson ,  Kristina Toutanova

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。

公开(公告)号：US20090106173A1

公开(公告)日：2009-04-23

申请号：US11874199

申请日：2007-10-17

申请人： Galen Andrew ,  Jianfeng Gao

发明人： Galen Andrew ,  Jianfeng Gao

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。

公开(公告)号：US07844555B2

公开(公告)日：2010-11-30

申请号：US11938811

申请日：2007-11-13

申请人： Jianfeng Gao ,  Galen Andrew ,  Mark Johnson ,  Kristina Toutanova

发明人： Jianfeng Gao ,  Galen Andrew ,  Mark Johnson ,  Kristina Toutanova

IPC分类号： G06F15/18 ,  G06E1/00 ,  G06E3/00 ,  G06G7/00 ,  G06N3/02

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。

公开(公告)号：US07984004B2

公开(公告)日：2011-07-19

申请号：US12015519

申请日：2008-01-17

申请人： Galen Andrew ,  Sooho Park ,  Robert L. Rounthwaite ,  Silviu-Petru Cucerzan ,  Jamie Paul Buckley ,  Joanna Chan

发明人： Galen Andrew ,  Sooho Park ,  Robert L. Rounthwaite ,  Silviu-Petru Cucerzan ,  Jamie Paul Buckley ,  Joanna Chan

IPC分类号： G06F17/00 ,  G06F7/00 ,  G06F17/30 ,  G06N5/00

CPC分类号： G06F17/30867 ,  G06F17/3064

摘要： Described herein is a system that facilitates assigning indications of usefulness to query suggestions. The system includes a query suggestion generator component that receives a query and generates a query suggestion based at least in part upon the received query. A model component outputs an indication of usefulness with respect to the query suggestion, wherein the model component is a machine-learned model of user behavior with respect to query suggestions.

摘要翻译： 这里描述了一种有助于分配对查询建议有用的指示的系统。 该系统包括查询建议生成器组件，其接收查询并且至少部分地基于所接收的查询来生成查询建议。 模型组件输出关于查询建议的有用性的指示，其中模型组件是关于查询建议的用户行为的机器学习模型。

公开(公告)号：US20100325133A1

公开(公告)日：2010-12-23

申请号：US12488603

申请日：2009-06-22

申请人： Robert L. Rounthwaite ,  Galen Andrew ,  Emre Mehmet Kiciman ,  Xiaoxin Yin

发明人： Robert L. Rounthwaite ,  Galen Andrew ,  Emre Mehmet Kiciman ,  Xiaoxin Yin

IPC分类号： G06F17/30

CPC分类号： G06F17/30864

摘要： A system described herein includes a receiver component that receives a dataset that is stored in a computer-readable medium of a computing device, wherein the dataset includes a plurality of queries issued by users to a search engine and a plurality of search results selected by the users upon issuing the plurality of queries. A distribution determiner component determines click distributions over the search results selected by the users with respect to the plurality of queries. A labeler component labels at least two queries in the plurality of queries as being substantially similar to one another based at least in part upon the click distributions over the search results selected by the users with respect to the plurality of queries.

摘要翻译： 本文描述的系统包括接收存储在计算设备的计算机可读介质中的数据集的接收器组件，其中所述数据集包括用户向搜索引擎发出的多个查询以及由所述搜索引擎选择的多个搜索结果 用户在发出多个查询时。 分布确定器组件确定用户相对于多个查询选择的搜索结果的点击分布。 标签器组件至少部分地基于用户相对于多个查询选择的搜索结果上的点击分布，将多个查询中的至少两个查询标记为彼此基本相似。

公开(公告)号：US20090187515A1

公开(公告)日：2009-07-23

申请号：US12015519

申请日：2008-01-17

申请人： Galen Andrew ,  Sooho Park ,  Robert L. Rounthwaite ,  Silviu-Petru Cucerzan ,  Jamie Paul Buckley ,  Joanna Chan

发明人： Galen Andrew ,  Sooho Park ,  Robert L. Rounthwaite ,  Silviu-Petru Cucerzan ,  Jamie Paul Buckley ,  Joanna Chan

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

CPC分类号： G06F17/30867 ,  G06F17/3064

摘要： Described herein is a system that facilitates assigning indications of usefulness to query suggestions. The system includes a query suggestion generator component that receives a query and generates a query suggestion based at least in part upon the received query. A model component outputs an indication of usefulness with respect to the query suggestion, wherein the model component is a machine-learned model of user behavior with respect to query suggestions.

摘要翻译： 这里描述了一种有助于分配对查询建议有用的指示的系统。 该系统包括查询建议生成器组件，其接收查询并且至少部分地基于所接收的查询来生成查询建议。 模型组件输出关于查询建议的有用性的指示，其中模型组件是关于查询建议的用户行为的机器学习模型。

公开(公告)号：US08606786B2

公开(公告)日：2013-12-10

申请号：US12488603

申请日：2009-06-22

申请人： Robert L. Rounthwaite ,  Galen Andrew ,  Emre Mehmet Kiciman ,  Xiaoxin Yin

发明人： Robert L. Rounthwaite ,  Galen Andrew ,  Emre Mehmet Kiciman ,  Xiaoxin Yin

IPC分类号： G06F7/00 ,  G06F17/30

CPC分类号： G06F17/30864

摘要： A system described herein includes a receiver component that receives a dataset that is stored in a computer-readable medium of a computing device, wherein the dataset includes a plurality of queries issued by users to a search engine and a plurality of search results selected by the users upon issuing the plurality of queries. A distribution determiner component determines click distributions over the search results selected by the users with respect to the plurality of queries. A labeler component labels at least two queries in the plurality of queries as being substantially similar to one another based at least in part upon the click distributions over the search results selected by the users with respect to the plurality of queries.

摘要翻译： 本文描述的系统包括接收存储在计算设备的计算机可读介质中的数据集的接收器组件，其中所述数据集包括用户向搜索引擎发出的多个查询以及由所述搜索引擎选择的多个搜索结果 用户在发出多个查询时。 分布确定器组件确定用户相对于多个查询选择的搜索结果的点击分布。 标签器组件至少部分地基于用户相对于多个查询选择的搜索结果上的点击分布，将多个查询中的至少两个查询标记为彼此基本相似。

公开(公告)号：US20070083357A1

公开(公告)日：2007-04-12

申请号：US11485015

申请日：2006-07-12

申请人： Robert Moore ,  Wen-tau Yih ,  Galen Andrew ,  Kristina Toutanova

发明人： Robert Moore ,  Wen-tau Yih ,  Galen Andrew ,  Kristina Toutanova

IPC分类号： G06F17/28

CPC分类号： G06F17/2827 ,  G06F17/2836

摘要： A weighted linear word alignment model linearly combines weighted features to score a word alignment for a bilingual, aligned pair of text fragments. The features are each weighted by a feature weight. One of the features is a word association metric, which may be generated from surface statistics.

摘要翻译： 加权线性字对齐模型线性组合加权特征以对双语对齐的文本片段对进行字对齐。 特征各自由特征权重加权。 特征之一是字关联度量，其可以从表面统计量生成。