公开(公告)号：US07231393B1

公开(公告)日：2007-06-12

申请号：US10788837

申请日：2004-02-26

申请人： Georges Harik ,  Noam M. Shazeer

发明人： Georges Harik ,  Noam M. Shazeer

IPC分类号： G06F17/30

CPC分类号： G06F17/30684 ,  G06F17/27

摘要： One embodiment of the present invention provides a system that learns a generative model for textual documents. During operation, the system receives a current model, which contains terminal nodes representing random variables for words and cluster nodes representing clusters of conceptually related words. Within the current model, nodes are coupled together by weighted links, so that if a cluster node in the probabilistic model fires, a weighted link from the cluster node to another node causes the other node to fire with a probability proportionate to the link weight. The system also receives a set of training documents, wherein each training document contains a set of words. Next, the system applies the set of training documents to the current model to produce a new model.

摘要翻译： 本发明的一个实施例提供一种学习文本文档的生成模型的系统。 在操作期间，系统接收当前模型，其包含表示用于表示概念相关词的群集的单词和群集节点的随机变量的终端节点。 在当前的模型中，节点通过加权链路耦合在一起，从而如果概率模型中的一个群集节点被触发，则从群集节点到另一个节点的加权链路会导致另一个节点以与链路权重成比例的概率发射。 该系统还接收一组训练文件，其中每个训练文档包含一组单词。 接下来，系统将该组培训文件应用于当前模型以产生新模型。

公开(公告)号：US07383258B2

公开(公告)日：2008-06-03

申请号：US10676571

申请日：2003-09-30

申请人： Georges Harik ,  Noam M. Shazeer

发明人： Georges Harik ,  Noam M. Shazeer

IPC分类号： G06F13/30

CPC分类号： G06F17/30707 ,  G06F17/30616 ,  G06F17/30705 ,  Y10S707/99935 ,  Y10S707/99943

摘要： One embodiment of the present invention provides a system characterizes a document with respect to clusters of conceptually related words. Upon receiving a document containing a set of words, the system selects “candidate clusters” of conceptually related words that are related to the set of words. These candidate clusters are selected using a model that explains how sets of words are generated from clusters of conceptually related words. Next, the system constructs a set of components to characterize the document, wherein the set of components includes components for candidate clusters. Each component in the set of components indicates a degree to which a corresponding candidate cluster is related to the set of words.

摘要翻译： 本发明的一个实施例提供了一种系统表征关于概念上相关词组的文档。 在接收到包含一组单词的文档时，系统选择与该组词相关的概念上相关词的“候选聚类”。 使用一个模型来选择这些候选群集，该模型解释了如何从概念上相关的单词群集中生成单词组。 接下来，系统构造一组组件来表征文档，其中该组组件包括用于候选聚类的组件。 组件组中的每个组件表示对应的候选集群与该组集合相关的程度。

公开(公告)号：US20150169507A1

公开(公告)日：2015-06-18

申请号：US13280826

申请日：2011-10-25

申请人： Noam M. Shazeer ,  Georges Harik

发明人： Noam M. Shazeer ,  Georges Harik

IPC分类号： G06F17/22 ,  G06F17/30

CPC分类号： G06F17/2235 ,  G06F17/30867

摘要： A method and an apparatus to provide a personalized page to a user have been disclosed. In one embodiment, a user is identified as a member of a first group and a member of a second group. The first group's level of interest (LOI) in a first item is identified, as well as the second group's LOI in a second item. The user's LOI in at least one of the first and the second items is identified.

摘要翻译： 已经公开了一种向用户提供个性化页面的方法和装置。 在一个实施例中，用户被识别为第一组的成员和第二组的成员。 第一组中的第一组兴趣水平（LOI）被确定，第二组的LOI在第二个项目中被识别。 识别在第一和第二项中的至少一个中的用户的LOI。

公开(公告)号：US08688720B1

公开(公告)日：2014-04-01

申请号：US12131637

申请日：2008-06-02

申请人： Georges Harik ,  Noam M. Shazeer

发明人： Georges Harik ,  Noam M. Shazeer

IPC分类号： G06F7/00

CPC分类号： G06F17/30707 ,  G06F17/30616 ,  G06F17/30705 ,  Y10S707/99935 ,  Y10S707/99943

摘要： One embodiment of the present invention provides a system characterizes a document with respect to clusters of conceptually related words. Upon receiving a document containing a set of words, the system selects “candidate clusters” of conceptually related words that are related to the set of words. These candidate clusters are selected using a model that explains how sets of words are generated from clusters of conceptually related words. Next, the system constructs a set of components to characterize the document, wherein the set of components includes components for candidate clusters. Each component in the set of components indicates a degree to which a corresponding candidate cluster is related to the set of words.

摘要翻译： 本发明的一个实施例提供了一种系统表征关于概念上相关词组的文档。 在接收到包含一组单词的文档时，系统选择与该组词相关的概念上相关词的“候选聚类”。 使用一个模型来选择这些候选群集，该模型解释了如何从概念上相关的单词群集中生成单词组。 接下来，系统构造一组组件来表征文档，其中该组组件包括用于候选聚类的组件。 组件组中的每个组件表示对应的候选集群与该组集合相关的程度。

公开(公告)号：US08412747B1

公开(公告)日：2013-04-02

申请号：US13237861

申请日：2011-09-20

申请人： Georges Harik ,  Noam M. Shazeer

发明人： Georges Harik ,  Noam M. Shazeer

IPC分类号： G06F17/30

CPC分类号： G06F17/30684 ,  G06F17/27

摘要： One embodiment of the present invention provides a system that learns a generative model for textual documents. During operation, the system receives a current model, which contains terminal nodes representing random variables for words and cluster nodes representing clusters of conceptually related words. Within the current model, nodes are coupled together by weighted links, so that if a cluster node in the probabilistic model fires, a weighted link from the cluster node to another node causes the other node to fire with a probability proportionate to the link weight. The system also receives a set of training documents, wherein each training document contains a set of words. Next, the system applies the set of training documents to the current model to produce a new model.

摘要翻译： 本发明的一个实施例提供一种学习文本文档的生成模型的系统。 在操作期间，系统接收当前模型，其包含表示用于表示概念相关词的群集的单词和群集节点的随机变量的终端节点。 在当前的模型中，节点通过加权链路耦合在一起，从而如果概率模型中的一个群集节点被触发，则从群集节点到另一个节点的加权链路会导致另一个节点以与链路权重成比例的概率发射。 该系统还接收一组训练文件，其中每个训练文档包含一组单词。 接下来，系统将该组培训文件应用于当前模型以产生新模型。

公开(公告)号：US09141589B2

公开(公告)日：2015-09-22

申请号：US13280826

申请日：2011-10-25

申请人： Noam M. Shazeer ,  Georges Harik

发明人： Noam M. Shazeer ,  Georges Harik

IPC分类号： G06F17/30 ,  G06F17/22

CPC分类号： G06F17/2235 ,  G06F17/30867

摘要： A method and an apparatus to provide a personalized page to a user have been disclosed. In one embodiment, a user is identified as a member of a first group and a member of a second group. The first group's level of interest (LOI) in a first item is identified, as well as the second group's LOI in a second item. The user's LOI in at least one of the first and the second items is identified.

摘要翻译： 已经公开了一种向用户提供个性化页面的方法和装置。 在一个实施例中，用户被识别为第一组的成员和第二组的成员。 第一组中的第一组兴趣水平（LOI）被确定，第二组的LOI在第二个项目中被识别。 识别在第一和第二项中的至少一个中的用户的LOI。

公开(公告)号：US08024372B2

公开(公告)日：2011-09-20

申请号：US11796383

申请日：2007-04-27

申请人： Georges Harik ,  Noam M. Shazeer

发明人： Georges Harik ,  Noam M. Shazeer

IPC分类号： G06F17/30

CPC分类号： G06F17/30684 ,  G06F17/27

摘要： One embodiment of the present invention provides a system that learns a generative model for textual documents. During operation, the system receives a current model, which contains terminal nodes representing random variables for words and cluster nodes representing clusters of conceptually related words. Within the current model, nodes are coupled together by weighted links, so that if a cluster node in the probabilistic model fires, a weighted link from the cluster node to another node causes the other node to fire with a probability proportionate to the link weight. The system also receives a set of training documents, wherein each training document contains a set of words. Next, the system applies the set of training documents to the current model to produce a new model.

摘要翻译： 本发明的一个实施例提供一种学习文本文档的生成模型的系统。 在操作期间，系统接收当前模型，其包含表示用于表示概念相关词的群集的单词和群集节点的随机变量的终端节点。 在当前的模型中，节点通过加权链路耦合在一起，从而如果概率模型中的一个群集节点被触发，则从群集节点到另一个节点的加权链路会导致另一个节点以与链路权重成比例的概率发射。 该系统还接收一组训练文件，其中每个训练文档包含一组单词。 接下来，系统将该组培训文件应用于当前模型以产生新模型。

公开(公告)号：US08661029B1

公开(公告)日：2014-02-25

申请号：US11556143

申请日：2006-11-02

申请人： Hyung-Jin Kim ,  Simon Tong ,  Noam M. Shazeer ,  Michelangelo Diligenti

发明人： Hyung-Jin Kim ,  Simon Tong ,  Noam M. Shazeer ,  Michelangelo Diligenti

IPC分类号： G06F7/00 ,  G06F17/30

CPC分类号： G06F17/3053 ,  G06F17/30696 ,  G06F17/30864 ,  G06F17/30867 ,  G06Q30/02

摘要： Systems and techniques relating to ranking search results of a search query include, in general, subject matter that can be embodied in a computer-implemented method that includes determining a measure of relevance for a document result within a context of a search query for which the document result is returned, the determining being based on a first number in relation to a second number, the first number corresponding to longer views of the document result, and the second number corresponding to at least shorter views of the document result; and outputting the measure of relevance to a ranking engine for ranking of search results, including the document result, for a new search corresponding to the search query. The subject matter described in this specification can also be embodied in various corresponding computer program products, apparatus and systems.

摘要翻译： 与搜索查询的搜索结果排序相关的系统和技术通常包括可以以计算机实现的方法体现的主题，包括确定在搜索查询的上下文内的文档结果的相关度的度量， 返回文档结果，所述确定基于与第二数字相关的第一数字，对应于文档结果的较长视图的第一数字和对应于文档结果的至少较短视图的第二数字; 并且对于与搜索查询相对应的新搜索，输出与包括文档结果在内的搜索结果排序相关的排名引擎的度量。 本说明书中描述的主题也可以体现在各种相应的计算机程序产品，装置和系统中。

公开(公告)号：US20100211894A1

公开(公告)日：2010-08-19

申请号：US12388245

申请日：2009-02-18

申请人： Michael E. Jahr ,  Uri N. Lerner ,  Noam M. Shazeer

发明人： Michael E. Jahr ,  Uri N. Lerner ,  Noam M. Shazeer

IPC分类号： G06F17/30 ,  G06F3/048

CPC分类号： G06F16/904

摘要： Among other disclosed subject matter, a computer-implemented method includes identifying a first object that belongs to a first domain. The method includes identifying, using the first object, at least a first cluster node in a generative model that includes a plurality of first cluster nodes having weighted relationships to respective ones of a plurality of second objects. The method includes identifying, in response to identifying the first object, at least one of the second objects, the second object belonging to the first domain and being identified using the first cluster node and its respective weighted relationship.

摘要翻译： 在其他公开的主题中，计算机实现的方法包括识别属于第一域的第一对象。 该方法包括使用第一对象来识别生成模型中的至少第一集群节点，其包括具有与多个第二对象中的相应的对象的加权关系的多个第一集群节点。 所述方法包括：响应于识别所述第一对象而识别所述第二对象中的至少一个，所述第二对象属于所述第一域并且使用所述第一集群节点及其相应的加权关系进行标识。