专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明授权

US08935258B2 Identification of sample data items for re-judging 有权
标题翻译：识别重新评估的样本数据项
公开(公告)号：US08935258B2
公开(公告)日：2015-01-13
申请号：US12484256
申请日：2009-06-15
申请人： Krysta M. Svore , Elbio Renato Torres Abib , Christopher J. C. Burges , Bhuvan Middha
发明人： Krysta M. Svore , Elbio Renato Torres Abib , Christopher J. C. Burges , Bhuvan Middha
IPC分类号： G06F17/30 , G06K9/62
CPC分类号： G06F17/30864 , G06K9/6257
摘要： Described is a technology for identifying sample data items (e.g., documents corresponding to query-URL pairs) having the greatest likelihood of being mislabeled when previously judged, and selecting those data items for re-judging. In one aspect, lambda gradient scores (information associated with ranked sample data items that indicates a relative direction and how “strongly” to move each data item for lowering a ranking cost) are summed for pairs of sample data items to compute re-judgment scores for each of those sample data items. The re-judgment scores indicate a relative likelihood of mislabeling. Once the selected sample data items are re-judged, a new training set is available, whereby a new ranker may be trained.
摘要翻译：描述了用于在先前判断时识别具有错误标记的最大可能性的样本数据项（例如，与查询 - URL对相对应的文档）的技术，并且选择那些用于重新判断的数据项。在一个方面，λ梯度分数（与指示相对方向的排名样本数据项相关联的信息以及如何“强烈地”移动每个数据项以降低排名成本）被求和于样本数据项对以计算重新判断分数对于每个样本数据项。重新判断分数表明错误标签的相对可能性。一旦选择的样本数据项被重新判断，就可以使用新的训练集，由此可以训练新的训练者。

2. 发明申请

US20100318540A1 IDENTIFICATION OF SAMPLE DATA ITEMS FOR RE-JUDGING 有权
标题翻译：鉴定样品数据项目进行重新评估
公开(公告)号：US20100318540A1
公开(公告)日：2010-12-16
申请号：US12484256
申请日：2009-06-15
申请人： Krysta M. Svore , Elbio Renato Torres Abib , Christopher J.C. Burges , Bhuvan Middha
发明人： Krysta M. Svore , Elbio Renato Torres Abib , Christopher J.C. Burges , Bhuvan Middha
IPC分类号： G06F17/30
CPC分类号： G06F17/30864 , G06K9/6257
摘要： Described is a technology for identifying sample data items (e.g., documents corresponding to query-URL pairs) having the greatest likelihood of being mislabeled when previously judged, and selecting those data items for re-judging. In one aspect, lambda gradient scores (information associated with ranked sample data items that indicates a relative direction and how “strongly” to move each data item for lowering a ranking cost) are summed for pairs of sample data items to compute re-judgment scores for each of those sample data items. The re-judgment scores indicate a relative likelihood of mislabeling. Once the selected sample data items are re-judged, a new training set is available, whereby a new ranker may be trained.
摘要翻译：描述了用于在先前判断时识别具有错误标记的最大可能性的样本数据项（例如，与查询 - URL对相对应的文档）的技术，并且选择那些用于重新判断的数据项。在一个方面，λ梯度分数（与指示相对方向的排名样本数据项相关联的信息以及如何“强烈地”移动每个数据项以降低排名成本）被求和于样本数据项对以计算重新判断分数对于每个样本数据项。重新判断分数表明错误标签的相对可能性。一旦选择的样本数据项被重新判断，就可以使用新的训练集，由此可以训练新的训练者。

3. 发明申请

US20130238608A1 SEARCH RESULTS BY MAPPING ASSOCIATED WITH DISPARATE TAXONOMIES 审中-公开
标题翻译：通过映射与不同的税收相关的搜索结果
公开(公告)号：US20130238608A1
公开(公告)日：2013-09-12
申请号：US13413651
申请日：2012-03-07
申请人： Ka Cheung Sia , Kyrylo Tropin , Bhuvan Middha , Paul Nathan Bennett , Krysta M. Svore
发明人： Ka Cheung Sia , Kyrylo Tropin , Bhuvan Middha , Paul Nathan Bennett , Krysta M. Svore
IPC分类号： G06F17/30
CPC分类号： G06F16/334
摘要： Architecture that generates signals/features that capture the match between intent of a query and category of documents. For example, for a query intent related to “autos”, documents that belong to categories related to “Autos” receive a higher score than documents of a “computers” category. The architecture can be applied to a search ecosystem where query intent classification and document category classifier are available, learns the mapping between query intent and document category, and introduces category-match features to a ranking algorithm, thereby improving search result relevance. The architecture learns the mapping between two existing and different taxonomies to create a category match signal from which the ranking algorithm can learn. Moreover, architecture adapts to a complex ecosystem where different taxonomies on the query side and document side exist through learning a mapping score between at least two taxonomies.
摘要翻译：生成能够捕获查询意图和文档类别之间匹配的信号/特征的体系结构。例如，对于与“autos”相关的查询意图，属于“Autos”相关类别的文档比“计算机”类别的文档获得更高的分数。该架构可以应用于查询意图分类和文档类别分类器可用的搜索生态系统，学习查询意图与文档类别之间的映射，并将类别匹配特征引入排序算法，从而提高搜索结果的相关性。该架构学习了两个现有和不同分类法之间的映射，以创建一个类别匹配信号，排序算法可从该类别匹配信号学习。此外，架构适应复杂的生态系统，通过学习至少两个分类法之间的映射分数，可以在查询方和文档方面存在不同的分类。

4. 发明授权

US09779359B2 Quantum arithmetic on two-dimensional quantum architectures 有权
公开(公告)号：US09779359B2
公开(公告)日：2017-10-03
申请号：US13420247
申请日：2012-03-14
申请人： Krysta M. Svore , Paul Tan The Pham
发明人： Krysta M. Svore , Paul Tan The Pham
IPC分类号： G06N99/00 , B82Y10/00 , G06F7/72
CPC分类号： G06N99/002 , B82Y10/00 , G06F7/72
摘要： 2D nearest-neighbor quantum architectures for Shor's factoring algorithm may be accomplished using the form of three arithmetic building blocks: modular addition using Gossett's carry-save addition, modular multiplication using Montgomery's method, and non-modular multiplication using an original method. These arithmetic building blocks may assume that ancillae are cheap, that concurrent control may be available and scalable, and that execution time may be the bottleneck. Thus, the arithmetic building blocks may be optimized in favor of circuit width to provide improved depth existing nearest-neighbor implementations.

5. 发明申请

US20130246495A1 Quantum Arithmetic On Two-Dimensional Quantum Architectures 有权
标题翻译：二维量子架构的量子算法
公开(公告)号：US20130246495A1
公开(公告)日：2013-09-19
申请号：US13420247
申请日：2012-03-14
申请人： Krysta M. Svore , Paul Tan The Pham
发明人： Krysta M. Svore , Paul Tan The Pham
IPC分类号： G06F7/501
CPC分类号： G06N99/002 , B82Y10/00 , G06F7/72
摘要： 2D nearest-neighbor quantum architectures for Shor's factoring algorithm may be accomplished using the form of three arithmetic building blocks: modular addition using Gossett's carry-save addition, modular multiplication using Montgomery's method, and non-modular multiplication using an original method. These arithmetic building blocks may assume that ancillae are cheap, that concurrent control may be available and scalable, and that execution time may be the bottleneck. Thus, the arithmetic building blocks may be optimized in favor of circuit width to provide improved depth existing nearest-neighbor implementations.
摘要翻译：用于Shor因子分解算法的2D最近邻量子架构可以使用三种算术构建模块的形式来实现：使用Gossett的进位保存加法的模块加法，使用蒙哥马利方法的模乘法和使用原始方法的非模乘法。这些算术构建块可以假设辅助系统是便宜的，并发控制可以是可用的和可扩展的，并且执行时间可能是瓶颈。因此，可以优化算术构建块以有利于电路宽度，以提供改进的深度存在的最近邻实现。

6. 发明申请

US20120158710A1 MULTI-TIERED INFORMATION RETRIEVAL TRAINING 审中-公开
标题翻译：多层次的信息检索培训
公开(公告)号：US20120158710A1
公开(公告)日：2012-06-21
申请号：US12974704
申请日：2010-12-21
申请人： Chris J.C. Burges , Krysta M. Svore , Maksims Volkovs
发明人： Chris J.C. Burges , Krysta M. Svore , Maksims Volkovs
IPC分类号： G06F17/30
CPC分类号： G06F16/3347 , G06F16/3326 , G06N20/00
摘要： Methods and systems for multi-tiered information retrieval training are disclosed. A method includes identifying results in a ranked ordering of results that can be swapped without changing a score determined using a first ranking quality measure, determining a first vector and at least one other vector for each identified swappable result in the ranked ordering of results based on the first ranking quality measure and at least one other ranking quality measure respectively, and adding the first vector and the at least one other vector for each identified swappable result in the ranked ordering of results to obtain a function of the first vector and the at least one other vector. Access is provided to the function of the first vector and the at least one other vector for use in the multi-tiered information retrieval training.
摘要翻译：披露了多层次信息检索培训的方法和系统。一种方法包括：在不改变使用第一排序质量度量确定的分数的情况下，将结果的排序顺序识别为结果排序的结果，确定第一向量和至少一个其他向量，以便根据排序的结果顺序，基于第一等级质量度量和至少一个其他排序质量度量，并且对于每个所识别的可交换结果，对结果的排序顺序添加第一向量和至少一个其他向量，以获得第一向量的功能，并且至少另一个向量。提供对第一向量和至少一个其他向量的功能的访问以用于多层次信息检索训练。

7. 发明授权

US10346453B2 Multi-tiered information retrieval training 有权
公开(公告)号：US10346453B2
公开(公告)日：2019-07-09
申请号：US12974704
申请日：2010-12-21
申请人： Chris J. C. Burges , Krysta M. Svore , Maksims Volkovs
发明人： Chris J. C. Burges , Krysta M. Svore , Maksims Volkovs
IPC分类号： G06F16/33 , G06F16/332 , G06N20/00
摘要： Methods and systems for multi-tiered information retrieval training are disclosed. A method includes identifying results in a ranked ordering of results that can be swapped without changing a score determined using a first ranking quality measure, determining a first vector and at least one other vector for each identified swappable result in the ranked ordering of results based on the first ranking quality measure and at least one other ranking quality measure respectively, and adding the first vector and the at least one other vector for each identified swappable result in the ranked ordering of results to obtain a function of the first vector and the at least one other vector. Access is provided to the function of the first vector and the at least one other vector for use in the multi-tiered information retrieval training.

8. 发明申请

US20130091128A1 Time-Aware Ranking Adapted to a Search Engine Application 有权
标题翻译：适应搜索引擎应用程序的时间意识排名
公开(公告)号：US20130091128A1
公开(公告)日：2013-04-11
申请号：US13270690
申请日：2011-10-11
申请人： Kira Radinsky , Susan T. Dumais , Krysta M. Svore , Jaime Brooks Teevan , Eric J. Horvitz
发明人： Kira Radinsky , Susan T. Dumais , Krysta M. Svore , Jaime Brooks Teevan , Eric J. Horvitz
IPC分类号： G06F17/30
CPC分类号： G06F17/3053 , G06F17/30129 , G06F17/30864 , G06F17/30887 , G06N5/045 , G06N99/005
摘要： Techniques provide time-aware ranking, such as ranking of information, files or URL (uniform resource locator) links. For example, time-aware modeling assists in determining user intent of a query to a search engine. In response to the query, results are ranked in a time-aware manner to better match the user intent. The ranking may model query, URL and query-URL pair behavior over time to create time-aware query, URL and query-URL pair models, respectively. Such models may predict behavior of a query-URL pair, such as frequency and timing of clicks to the URL of the pair when the query of the pair is posed to the search engine. Results of a query may be ranked by predicted query-URL behavior. Once ranked, the results may be sent to the user in response to the query.
摘要翻译：技术提供时间感知排名，例如信息排名，文件或URL（统一资源定位符）链接。例如，时间感知建模有助于确定对搜索引擎的查询的用户意图。响应于查询，结果以时间感知的方式排列，以更好地匹配用户意图。排名可能会对查询，URL和查询 - URL对行为进行模拟，以分别创建时间感知查询，URL和查询 - URL对模型。这样的模型可以预测查询 - URL对的行为，例如当该对的查询提供给搜索引擎时，对对方的URL的点击的频率和时间。查询结果可能会按预测的查询网址行为进行排名。一旦排序，结果可能会发送给用户以响应查询。

9. 发明申请

US20110208735A1 Learning Term Weights from the Query Click Field for Web Search 审中-公开
标题翻译：从查询的学习术语权重单击字段进行Web搜索
公开(公告)号：US20110208735A1
公开(公告)日：2011-08-25
申请号：US12710360
申请日：2010-02-23
申请人： Jianfeng Gao , Krysta M. Svore
发明人： Jianfeng Gao , Krysta M. Svore
IPC分类号： G06F15/18 , G06F17/30
CPC分类号： G06F16/951
摘要： Described is a technology by which a term frequency function for web click data is machine learned from raw click features extracted from a query log or the like and training data. Also described is using combining the term frequency function with other functions/click features to learn a relevance function for use in ranking document relevance to a query.
摘要翻译：描述了一种技术，通过该技术，从查询日志等提取的原始点击功能和训练数据中的机器学习了用于web点击数据的术语频率函数。还描述了将术语频率函数与其他功能/点击功能相结合，以学习用于对与查询进行文档相关性的相关性功能。

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