专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

11. 发明申请

US20060277173A1 Extraction of information from documents 有权
标题翻译：从文件中提取信息
公开(公告)号：US20060277173A1
公开(公告)日：2006-12-07
申请号：US11192687
申请日：2005-07-29
申请人： Hang Li , Ruihua Song , Yunbo Cao , Dmitriy Meyerzon
发明人： Hang Li , Ruihua Song , Yunbo Cao , Dmitriy Meyerzon
IPC分类号： G06F17/30 , G06F17/00 , G06F17/27
CPC分类号： G06F17/211 , Y10S707/99935
摘要： An information extraction model is trained on format features identified within labeled training documents. Information from a document is extracted by assigning labels to units based on format features of the units within the document. A begin label and end label are identified and the information is extracted between the begin label and the end label. The extracted information can be used in various document processing tasks such as ranking.
摘要翻译：对标示的培训文件中标识的格式特征进行信息提取模型的培训。通过根据文档中单位的格式特征为单位分配标签来提取文档中的信息。识别开始标签和结束标签，并在开始标签和结束标签之间提取信息。提取的信息可以用于各种文档处理任务，如排名。

12. 发明授权

US08898156B2 Query expansion for web search 有权
标题翻译：网页搜索的查询扩展
公开(公告)号：US08898156B2
公开(公告)日：2014-11-25
申请号：US13040192
申请日：2011-03-03
申请人： Jun Xu , Hang Li
发明人： Jun Xu , Hang Li
IPC分类号： G06F17/30
CPC分类号： G06F17/30864 , G06F17/30672
摘要： Systems, methods, and devices are described for retrieving query results based at least in part on a query and one or more similar queries. Upon receiving a query, one or more similar queries may be identified and/or calculated. In one embodiment, the similar queries may be determined based at least in part on click-through data corresponding to previously submitted queries. Information associated with the query and each of the similar queries may be retrieved, ranked, and or combined. The combined query results may then be re-ranked based at least in part on a responsiveness and/or relevance to the previously submitted query. The re-ranked query results may then be output to a user that submitted the original query.
摘要翻译：描述了至少部分地基于查询和一个或多个类似查询来检索查询结果的系统，方法和设备。在接收到查询时，可以识别和/或计算一个或多个类似的查询。在一个实施例中，可以至少部分地基于对应于先前提交的查询的点击数据来确定类似的查询。与查询相关联的信息和每个相似查询可以被检索，排序和/或组合。组合的查询结果可以至少部分地基于对先前提交的查询的响应性和/或相关性来重新排序。然后可以将重新排列的查询结果输出给提交原始查询的用户。

13. 发明授权

US08533195B2 Regularized latent semantic indexing for topic modeling 有权
标题翻译：主题建模的正则化潜在语义索引
公开(公告)号：US08533195B2
公开(公告)日：2013-09-10
申请号：US13169808
申请日：2011-06-27
申请人： Jun Xu , Hang Li , Nicholas Craswell
发明人： Jun Xu , Hang Li , Nicholas Craswell
IPC分类号： G06F7/00 , G06F17/30 , G06F17/16 , G06F17/11 , G06F11/34
CPC分类号： G06F17/16 , G06F11/3447 , G06F17/11 , G06F17/30705 , G06F2212/454 , G06K9/00979 , G06K9/6249
摘要： Electronic documents are retrieved from a database and/or from a network of servers. The documents are topic modeled in accordance with a Regularized Latent Semantic Indexing approach. The Regularized Latent Semantic Indexing approach may allow an equation involving an approximation of a term-document matrix to be solved in parallel by multiple calculating units. The equation may include terms that are regularized via either l1 norm and/or via l2 norm. The Regularized Latent Semantic Indexing approach may be applied to a set, or a fixed number, of documents such that the set of documents is topic modeled. Alternatively, the Regularized Latent Semantic Indexing approach may be applied to a variable number of documents such that, over time, the variable of number of documents is topic modeled.
摘要翻译：从数据库和/或从服务器网络检索电子文档。这些文件是根据正则潜在语义索引方法建模的主题。正则潜在语义索引方法可以允许涉及术语文档矩阵的近似的等式由多个计算单元并行求解。方程式可以包括通过l1范数和/或通过l2范数规则化的项。正则潜在语义索引方法可以应用于一组或固定数量的文档，使得该组文档被主题建模。或者，正则潜在语义索引方法可以应用于可变数量的文档，使得随着时间的推移，文档数量的变量被主题建模。

14. 发明授权

US08478748B2 Directly optimizing evaluation measures in learning to rank 有权
标题翻译：直接优化学习排名评估指标
公开(公告)号：US08478748B2
公开(公告)日：2013-07-02
申请号：US12237293
申请日：2008-09-24
申请人： Jun Xu , Tie-Yan Liu , Hang Li
发明人： Jun Xu , Tie-Yan Liu , Hang Li
IPC分类号： G06F17/30
CPC分类号： G06F17/30687 , G06F17/30867
摘要： The present invention provides methods for improving a ranking model. In one embodiment, a method includes the step of obtaining queries, documents, and document labels. The process then initializes active sets using the document labels, wherein two active sets are established for each query, a perfect active set and an imperfect active set. Then, the process optimizes an empirical loss function by the use of the first and second active set, whereby parameters of the ranking model are modified in accordance to the empirical loss function. The method then updates the active sets with additional ranking data, wherein the updates are configured to work in conjunction with the optimized loss function and modified ranking model. The recalculated active sets provide an indication for ranking the documents in a way that is more consistent with the document metadata.
摘要翻译：本发明提供了改进排名模型的方法。在一个实施例中，一种方法包括获得查询，文档和文档标签的步骤。然后，该过程使用文档标签来初始化活动集合，其中为每个查询建立两个活动集合，完美的活动集合和不完全的活动集合。然后，该过程通过使用第一和第二活动集来优化经验损失函数，由此根据经验损失函数修改排名模型的参数。然后，该方法用附加排名数据更新活动集合，其中更新被配置为与优化的损失函数和修改的排名模型一起工作。重新计算的活动集提供了以与文档元数据更一致的方式对文档进行排名的指示。

15. 发明申请

US20120030200A1 TOPICS IN RELEVANCE RANKING MODEL FOR WEB SEARCH 有权
标题翻译：用于网络搜索的相关排名模式的主题
公开(公告)号：US20120030200A1
公开(公告)日：2012-02-02
申请号：US13271638
申请日：2011-10-12
申请人： Qing Yu , Jun Xu , Hang Li
发明人： Qing Yu , Jun Xu , Hang Li
IPC分类号： G06F17/30
CPC分类号： G06F17/30864
摘要： Described is a technology by which topics corresponding to web pages are used in relevance ranking of those pages. Topics are extracted from each web page of a set of web pages that are found via a query. For example, text such as nouns may be extracted from the title, anchor texts and URL of a page, and used as the topics. The extracted topics from a page are used to compute a relevance score for that page based on an evaluation of that page's topics against the query. The pages are then ranked relative to one another based at least in part on the relevance score computed for each page, such as by determining a matching level for each page, ranking pages by each level, and ranking pages within each level. Also described is training a model to perform the relevance scoring and/or ranking.
摘要翻译：描述了一种技术，通过该技术将与网页相对应的主题用于那些页面的相关性排名。从通过查询找到的一组网页的每个网页中提取主题。例如，可以从标题，锚文本和页面的URL中提取诸如名词的文本，并且用作主题。从页面提取的主题用于根据对该页面的主题对查询的评估来计算该页面的相关性分数。这些页面至少部分地基于针对每个页面计算的相关性分数相对于彼此进行排名，例如通过确定每个页面的匹配级别，按各级别排序页面以及在每个级别内对页面进行排序。还描述了训练模型以执行相关性评分和/或排名。

16. 发明申请

US20090327264A1 Topics in Relevance Ranking Model for Web Search 有权
标题翻译：网页搜索相关性排名模型的主题
公开(公告)号：US20090327264A1
公开(公告)日：2009-12-31
申请号：US12146430
申请日：2008-06-25
申请人： Qing Yu , Jun Xu , Hang Li
发明人： Qing Yu , Jun Xu , Hang Li
IPC分类号： G06F17/30 , G06F15/18
CPC分类号： G06F17/30864
摘要： Described is a technology by which topics corresponding to web pages are used in relevance ranking of those pages. Topics are extracted from each web page of a set of web pages that are found via a query. For example, text such as nouns may be extracted from the title, anchor texts and URL of a page, and used as the topics. The extracted topics from a page are used to compute a relevance score for that page based on an evaluation of that page's topics against the query. The pages are then ranked relative to one another based at least in part on the relevance score computed for each page, such as by determining a matching level for each page, ranking pages by each level, and ranking pages within each level. Also described is training a model to perform the relevance scoring and/or ranking.
摘要翻译：描述了一种技术，通过该技术将与网页相对应的主题用于那些页面的相关性排名。从通过查询找到的一组网页的每个网页中提取主题。例如，可以从标题，锚文本和页面的URL中提取诸如名词的文本，并且用作主题。从页面提取的主题用于根据对该页面的主题对查询的评估来计算该页面的相关性分数。这些页面至少部分地基于针对每个页面计算的相关性分数相对于彼此进行排名，例如通过确定每个页面的匹配级别，按各级别排序页面以及在每个级别内对页面进行排序。还描述了训练模型以执行相关性评分和/或排名。

17. 发明授权

US08065310B2 Topics in relevance ranking model for web search 有权
标题翻译：网络搜索的相关性排名模型中的主题
公开(公告)号：US08065310B2
公开(公告)日：2011-11-22
申请号：US12146430
申请日：2008-06-25
申请人： Qing Yu , Jun Xu , Hang Li
发明人： Qing Yu , Jun Xu , Hang Li
IPC分类号： G06F17/30
CPC分类号： G06F17/30864
摘要： Described is a technology by which topics corresponding to web pages are used in relevance ranking of those pages. Topics are extracted from each web page of a set of web pages that are found via a query. For example, text such as nouns may be extracted from the title, anchor texts and URL of a page, and used as the topics. The extracted topics from a page are used to compute a relevance score for that page based on an evaluation of that page's topics against the query. The pages are then ranked relative to one another based at least in part on the relevance score computed for each page, such as by determining a matching level for each page, ranking pages by each level, and ranking pages within each level. Also described is training a model to perform the relevance scoring and/or ranking.
摘要翻译：描述了一种技术，通过该技术将与网页相对应的主题用于那些页面的相关性排名。从通过查询找到的一组网页的每个网页中提取主题。例如，可以从标题，锚文本和页面的URL中提取诸如名词的文本，并且用作主题。从页面提取的主题用于根据对该页面的主题对查询的评估来计算该页面的相关性分数。这些页面至少部分地基于针对每个页面计算的相关性分数相对于彼此进行排名，例如通过确定每个页面的匹配级别，按各级别排序页面以及在每个级别内对页面进行排序。还描述了训练模型以执行相关性评分和/或排名。

18. 发明授权

US07877383B2 Ranking and accessing definitions of terms 失效
标题翻译：排名和访问术语的定义
公开(公告)号：US07877383B2
公开(公告)日：2011-01-25
申请号：US11115500
申请日：2005-04-27
申请人： Yunbo Cao , Hang Li , Jun Xu
发明人： Yunbo Cao , Hang Li , Jun Xu
IPC分类号： G06F7/00
CPC分类号： G06F17/30654 , G06F2216/03
摘要： A method of processing information is provided. The method includes collecting text strings of definition candidates from a data source. The definition candidates are ranked based on the text strings.
摘要翻译：提供了处理信息的方法。该方法包括从数据源收集定义候选的文本串。定义候选人基于文本字符串进行排名。

19. 发明申请

US20100082606A1 DIRECTLY OPTIMIZING EVALUATION MEASURES IN LEARNING TO RANK 有权
标题翻译：直接优化评估评估方法
公开(公告)号：US20100082606A1
公开(公告)日：2010-04-01
申请号：US12237293
申请日：2008-09-24
申请人： Jun Xu , Tie-Yan Liu , Hang Li
发明人： Jun Xu , Tie-Yan Liu , Hang Li
IPC分类号： G06F17/30 , G06F17/10
CPC分类号： G06F17/30687 , G06F17/30867
摘要： The present invention provides methods for improving a ranking model. In one embodiment, a method includes the step of obtaining queries, documents, and document labels. The process then initializes active sets using the document labels, wherein two active sets are established for each query, a perfect active set and an imperfect active set. Then, the process optimizes an empirical loss function by the use of the first and second active set, whereby parameters of the ranking model are modified in accordance to the empirical loss function. The method then updates the active sets with additional ranking data, wherein the updates are configured to work in conjunction with the optimized loss function and modified ranking model. The recalculated active sets provide an indication for ranking the documents in a way that is more consistent with the document metadata.
摘要翻译：本发明提供了改进排名模型的方法。在一个实施例中，一种方法包括获得查询，文档和文档标签的步骤。然后，该过程使用文档标签来初始化活动集合，其中为每个查询建立两个活动集合，完美的活动集合和不完全的活动集合。然后，该过程通过使用第一和第二活动集来优化经验损失函数，由此根据经验损失函数修改排名模型的参数。然后，该方法用附加排名数据更新活动集合，其中更新被配置为与优化的损失函数和修改的排名模型一起工作。重新计算的活动集提供了以与文档元数据更一致的方式对文档进行排名的指示。

20. 发明授权

US07593934B2 Learning a document ranking using a loss function with a rank pair or a query parameter 有权
标题翻译：使用具有排名对或查询参数的损失函数学习文档排名
公开(公告)号：US07593934B2
公开(公告)日：2009-09-22
申请号：US11460838
申请日：2006-07-28
申请人： Hang Li , Jun Xu , Yunbo Cao , Tie-Yan Liu
发明人： Hang Li , Jun Xu , Yunbo Cao , Tie-Yan Liu
IPC分类号： G06F7/00 , G06F17/30 , G06F15/00
CPC分类号： G06F17/30864 , Y10S707/99932 , Y10S707/99933 , Y10S707/99935 , Y10S707/99938 , Y10S707/99939
摘要： A method and system for generating a ranking function to rank the relevance of documents to a query is provided. The ranking system learns a ranking function from training data that includes queries, resultant documents, and relevance of each document to its query. The ranking system learns a ranking function using the training data by weighting incorrect rankings of relevant documents more heavily than the incorrect rankings of not relevant documents so that more emphasis is placed on correctly ranking relevant documents. The ranking system may also learn a ranking function using the training data by normalizing the contribution of each query to the ranking function so that it is independent of the number of relevant documents of each query.
摘要翻译：提供了一种用于生成用于将文档与查询的相关性排序的排序函数的方法和系统。排名系统从包括查询，结果文档以及每个文档与其查询的相关性的训练数据中学习排名函数。排名系统使用训练数据通过对相关文件的不正确排名加权比不相关文件的不正确排名更多地学习排名功能，以便更加重视正确排列相关文件。排序系统还可以通过将每个查询的贡献归一化到排序函数来学习使用训练数据的排序函数，使得它独立于每个查询的相关文档的数量。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式