专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

31. 发明授权

US08713014B1 Simplifying lexicon creation in hybrid duplicate detection and inductive classifier systems 有权
标题翻译：简化混合重复检测和归纳分类系统中的词典创建
公开(公告)号：US08713014B1
公开(公告)日：2014-04-29
申请号：US12780500
申请日：2010-05-14
申请人： Joshua Alspector , Abdur R. Chowdhury , Aleksander Kolcz
发明人： Joshua Alspector , Abdur R. Chowdhury , Aleksander Kolcz
IPC分类号： G06F17/30
CPC分类号： G06F17/30705 , G06F17/30156 , G06F17/30707 , G06N5/02 , H04L51/12
摘要： A classification system includes a signature-based duplicate detector and an inductive classifier that share attribute information. To perform the duplicate detection and the classification, the duplicate detector and inductive classifier are first initialized by generating a lexicon of attributes for the duplicate detector and a classification model for the classifier. To develop a classification model, a training set of documents of known class are used by the classifier to determine the attributes of the documents that are most useful in classifying an unknown document. The model is developed from these attributes. Attribute information containing the attributes determined by the classifier is then passed to the duplicate detector and the duplicate detector uses the attribute information to generate the lexicon of attributes.
摘要翻译：分类系统包括共享属性信息的基于签名的重复检测器和归纳分类器。为了执行重复检测和分类，首先通过生成用于重复检测器的属性的词典和分类器的分类模型来初始化重复检测器和归纳分类器。为了开发分类模型，分类器使用已知类的文档的训练集来确定在分类未知文档中最有用的文档的属性。该模型是从这些属性开发的。包含由分类器确定的属性的属性信息然后传递给重复检测器，并且重复检测器使用属性信息来生成属性的词典。

32. 发明申请

US20130173562A1 Simplifying Lexicon Creation in Hybrid Duplicate Detection and Inductive Classifier System 有权
标题翻译：简化混合重复检测和感应分类系统中的词典创建
公开(公告)号：US20130173562A1
公开(公告)日：2013-07-04
申请号：US13612841
申请日：2012-09-13
申请人： Joshua ALSPECTOR , Aleksander KOLCZ , Abdur R. CHOWDHURY
发明人： Joshua ALSPECTOR , Aleksander KOLCZ , Abdur R. CHOWDHURY
IPC分类号： G06F17/30
CPC分类号： G06F17/30705 , G06F17/30156 , G06F17/30707 , G06N5/02 , H04L51/12
摘要： A classification system includes a signature-based duplicate detector and an inductive classifier that share attribute information. To perform the duplicate detection and the classification, the duplicate detector and inductive classifier are first initialized by generating a lexicon of attributes for the duplicate detector and a classification model for the classifier. To develop a classification model, a training set of documents of known class are used by the classifier to determine the attributes of the documents that are most useful in classifying an unknown document. The model is developed from these attributes. Attribute information containing the attributes determined by the classifier is then passed to the duplicate detector and the duplicate detector uses the attribute information to generate the lexicon of attributes.
摘要翻译：分类系统包括共享属性信息的基于签名的重复检测器和归纳分类器。为了执行重复检测和分类，首先通过生成用于重复检测器的属性的词典和分类器的分类模型来初始化重复检测器和归纳分类器。为了开发分类模型，分类器使用已知类的文档的训练集来确定在分类未知文档中最有用的文档的属性。该模型是从这些属性开发的。包含由分类器确定的属性的属性信息然后传递给重复检测器，并且重复检测器使用属性信息来生成属性的词典。

33. 发明申请

US20120173560A1 QUERY ROUTING 审中-公开
标题翻译：查询路由
公开(公告)号：US20120173560A1
公开(公告)日：2012-07-05
申请号：US13418312
申请日：2012-03-12
申请人： Abdur R. CHOWDHURY , Gregory S. Pass
发明人： Abdur R. CHOWDHURY , Gregory S. Pass
IPC分类号： G06F17/30
CPC分类号： G06F16/954 , G06F16/334 , Y10S707/99933 , Y10S707/99934 , Y10S707/99935
摘要： A search query is submitted to one or more information sources associated with a category of the query. The category of the query is indicated by a query ontology that relates queries to query categories. The information sources represent information sources from which a high percentage of search results for queries associated with the category are retrieved. For instance, the category of the query is identified by identifying categories corresponding to variations of the query, where each variation represents a combination of the terms within the query, and where the categories of the variations are assumed to be the categories of the query. Information sources associated with the query categories are identified, and the query is submitted to the identified information sources. Submitting the query to the identified information sources may cause search results retrieved for the query to more closely reflect search results desired by a user that specified the query.
摘要翻译：搜索查询被提交给与查询的类别相关联的一个或多个信息源。查询的类别由查询本体指示，将查询与查询类别相关联。信息来源代表了与该类别相关联的查询的高比例搜索结果的信息源。例如，通过识别与查询的变体相对应的类别来识别查询的类别，其中每个变体表示查询中的术语的组合，以及哪些变体的类别被假设为查询的类别。识别与查询类别相关联的信息源，并将查询提交给标识的信息源。将查询提交到标识的信息源可能会导致为查询检索的搜索结果更加紧密地反映指定查询的用户所期望的搜索结果。

34. 发明申请

US20120150907A1 AUDIO AND/OR VIDEO SCENE DETECTION AND RETRIEVAL 有权
标题翻译：音频和/或视频场景检测和检索
公开(公告)号：US20120150907A1
公开(公告)日：2012-06-14
申请号：US13404610
申请日：2012-02-24
申请人： Abdur R. CHOWDHURY , Harmannus Vandermolen
发明人： Abdur R. CHOWDHURY , Harmannus Vandermolen
IPC分类号： G06F17/30
CPC分类号： G06F17/30843 , G06F17/30056 , G06F17/30787 , G11B27/002 , H04N21/23418 , H04N21/8549
摘要： Movie video trailers for a particular movie quote may be created and provided to a user. The Internet may be searched to identify documents that likely include references to a movie. A reference to the movie within an identified document may be detected and determined to be a movie quote. The movie quote and related information may be extracted from the identified document. A location of the movie quote within the movie may be determined. A movie video trailer that includes the movie quote may be created based on the location of the movie quote. A request for a movie video trailer that includes a movie quote or a partial movie quote, specified by the user, may be received from the user. A movie video trailer that includes the movie quote or the partial movie quote may be identified and provided to the user.
摘要翻译：可以创建特定电影报价的电影视频预告片并提供给用户。可以搜索互联网以识别可能包括对电影的引用的文档。可以检测并确定对所识别的文档内的电影的引用是电影报价。可以从所识别的文档中提取电影报价和相关信息。可以确定电影中的电影报价的位置。可以根据电影报价的位置创建包含电影报价的电影视频预告片。可以从用户接收到包括由用户指定的电影报价或部分电影报价的电影视频预告片的请求。可以识别包括电影报价或部分电影报价的电影视频预告片并将其提供给用户。

35. 发明授权

US08005813B2 Domain expert search 失效
标题翻译：域专家搜索
公开(公告)号：US08005813B2
公开(公告)日：2011-08-23
申请号：US11853877
申请日：2007-09-12
申请人： Abdur R. Chowdhury , Gregory S. Pass , Gerald Frederick Campbell
发明人： Abdur R. Chowdhury , Gregory S. Pass , Gerald Frederick Campbell
IPC分类号： G06F17/30 , G06F7/00
CPC分类号： G06F17/30864 , G06F17/30675 , G06F17/30876 , Y10S707/99933
摘要： Expert domains for a query category represent domains from which a high percentage of search results for queries associated with the query category are retrieved. The expert domains are identified by establishing a base statistical model that indicates frequencies of appearance for domains in search results retrieved for queries corresponding to multiple categories. In addition, frequencies of domain appearance are determined for search results retrieved for queries associated with a category. Domains that appear more frequently in the search results corresponding to the category are identified as expert domains for the category. A user may be allowed to customize expert domains related to one or more categories by adding or removing expert domains for the category.
摘要翻译：查询类别的专家域表示检索与查询类别相关联的查询的高比例搜索结果的域。通过建立基准统计模型来识别专家领域，该基础统计模型指示针对与多个类别相对应的查询检索的搜索结果中的域的出现频率。另外，针对与类别相关联的查询检索的搜索结果确定域外观的频率。在与该类别相对应的搜索结果中更频繁出现的域被标识为该类别的专家域。可以允许用户通过添加或删除该类别的专家域来定制与一个或多个类别相关的专家域。

36. 发明授权

US07984029B2 Reliability of duplicate document detection algorithms 有权
标题翻译：重复文件检测算法的可靠性
公开(公告)号：US07984029B2
公开(公告)日：2011-07-19
申请号：US12144021
申请日：2008-06-23
申请人： Joshua Alspector , Aleksander Kolcz , Abdur R. Chowdhury
发明人： Joshua Alspector , Aleksander Kolcz , Abdur R. Chowdhury
IPC分类号： G06F7/00
CPC分类号： G06F17/30156 , G06F17/30011 , Y10S707/99943 , Y10S707/99945
摘要： In a single-signature duplicate document system, a secondary set of attributes is used in addition to a primary set of attributes so as to improve the precision of the system. When the projection of a document onto the primary set of attributes is below a threshold, then a secondary set of attributes is used to supplement the primary lexicon so that the projection is above the threshold.
摘要翻译：在单签名重复文档系统中，除了主要属性集之外还使用辅助属性集，以提高系统的精度。当将文档投影到主要属性集合上时，使用辅助的一组属性来补充主要词典，使得投影高于阈值。

37. 发明授权

US07818314B2 Search fusion 有权
标题翻译：搜索融合
公开(公告)号：US07818314B2
公开(公告)日：2010-10-19
申请号：US11023642
申请日：2004-12-29
申请人： Abdur R. Chowdhury , Gregory S. Pass
发明人： Abdur R. Chowdhury , Gregory S. Pass
IPC分类号： G06F7/00 , G06F17/30
CPC分类号： G06F17/30864 , G06F17/30696
摘要： Search results are assigned scores based on visual aspects of surrogate representations of the search results which are to be displayed to a query submitter in a search results summary overview. That is, the surrogate representations are relatively short summaries or excerpts of the search results that may be presented in place of the search results themselves, thus enabling an overview of various search results to be perceived by a user concurrently. The search results are assigned scores based on the presence, position, and form of some or all of the query within the corresponding surrogate representations. The search results may be sorted or filtered based on the assigned scores. Assigning scores to the search results based on visual characteristics of the surrogate representations mimics how a user may assess the relevance of the search results when viewing a search results summary page.
摘要翻译：搜索结果基于搜索结果的代理表示的视觉方面的分配，这些搜索结果将在搜索结果摘要概述中显示给查询提交者。也就是说，替代表示是搜索结果的相对较短的摘要或摘录，可以代替搜索结果本身来呈现，从而使得用户可以同时感知各种搜索结果的概述。搜索结果根据相应替代表示中的部分或全部查询的存在，位置和形式进行分配。搜索结果可以根据分配的分数进行排序或过滤。基于代理表示的视觉特征为搜索结果分配分数模仿用户在查看搜索结果摘要页面时如何评估搜索结果的相关性。

38. 发明授权

US07779009B2 Web query classification 有权
标题翻译： Web查询分类
公开(公告)号：US07779009B2
公开(公告)日：2010-08-17
申请号：US11340843
申请日：2006-01-27
申请人： Abdur R. Chowdhury , Steven Michael Beitzel , David Dolan Lewis , Aleksander Kolcz
发明人： Abdur R. Chowdhury , Steven Michael Beitzel , David Dolan Lewis , Aleksander Kolcz
IPC分类号： G06F17/30
CPC分类号： G06F17/30707 , G06F17/30657 , G06F17/30864
摘要： A query phrase may be automatically classified to one or more topics of interest (e.g., categories) to assist in routing the query phrase to one or more appropriate backend databases. A selectional preference query classification technique may be used to classify the query phrase based on a comparison between the query phrase and patterns of query phrases. Additionally, or alternatively, a combination of query classification techniques may be used to classify the query phrase. Topical classification of a query phrase also may be used to assist a search system in delivering auxiliary information to a user who entered the query phrase. Advertisements, for instance, may be tailored based on classification rather than query keywords.
摘要翻译：查询短语可以被自动分类为感兴趣的一个或多个主题（例如，类别），以帮助将查询短语路由到一个或多个适当的后端数据库。可以使用选择偏好查询分类技术来基于查询短语和查询短语的模式之间的比较来对查询短语进行分类。另外或替代地，可以使用查询分类技术的组合来对查询短语进行分类。查询短语的主题分类也可以用于帮助搜索系统向输入查询短语的用户传送辅助信息。例如，广告可以基于分类来定制，而不是查询关键字。

39. 发明授权

US07725475B1 Simplifying lexicon creation in hybrid duplicate detection and inductive classifier systems 有权
标题翻译：简化混合重复检测和归纳分类系统中的词典创建
公开(公告)号：US07725475B1
公开(公告)日：2010-05-25
申请号：US11016930
申请日：2004-12-21
申请人： Joshua Alspector , Aleksander Kolcz , Abdur R. Chowdhury
发明人： Joshua Alspector , Aleksander Kolcz , Abdur R. Chowdhury
IPC分类号： G06F7/00
CPC分类号： G06F17/30705 , G06F17/30156 , G06F17/30707 , G06N5/02 , H04L51/12
摘要： A classification system includes a signature-based duplicate detector and an inductive classifier that share attribute information. To perform the duplicate detection and the classification, the duplicate detector and inductive classifier are first initialized by generating a lexicon of attributes for the duplicate detector and a classification model for the classifier. To develop a classification model, a training set of documents of known class are used by the classifier to determine the attributes of the documents that are most useful in classifying an unknown document. The model is developed from these attributes. Attribute information containing the attributes determined by the classifier is then passed to the duplicate detector and the duplicate detector uses the attribute information to generate the lexicon of attributes.
摘要翻译：分类系统包括共享属性信息的基于签名的重复检测器和归纳分类器。为了执行重复检测和分类，首先通过生成用于重复检测器的属性的词典和分类器的分类模型来初始化重复检测器和归纳分类器。为了开发分类模型，分类器使用已知类的文档的训练集来确定在分类未知文档中最有用的文档的属性。该模型是从这些属性开发的。包含由分类器确定的属性的属性信息然后传递给重复检测器，并且重复检测器使用属性信息来生成属性的词典。

40. 发明授权

US07689540B2 Collaborative user query refinement 有权
标题翻译：协作用户查询优化
公开(公告)号：US07689540B2
公开(公告)日：2010-03-30
申请号：US11613004
申请日：2006-12-19
申请人： Abdur R. Chowdhury , Harmannus Vandermolen , Eric Jensen
发明人： Abdur R. Chowdhury , Harmannus Vandermolen , Eric Jensen
IPC分类号： G06F7/00
CPC分类号： G06F17/30867 , G06F17/30648
摘要： A first query and a second query, both of which were submitted by a first user to identify information that satisfies a single information need of the first user, are received. The first query and the second query are stored in association with one another and based on a chronological order in which the first query and the second query were submitted by the first user to generate a first query reformulation path. A query is received from a user and more than one query reformulation path is accessed. The received query is compared with queries included in the accessed query reformulation paths to identify a relationship between the received query and a query included in a relevant query reformulation path. One or more prevalent end points is determined based on the relevant query reformulation path and at least one of the prevalent end points is presented to the user.
摘要翻译：接收由第一用户提交的用于识别满足第一用户的单个信息需求的信息的第一查询和第二查询。第一查询和第二查询相互关联地存储，并且基于第一用户提交的第一查询和第二查询的时间顺序以生成第一查询重写路径。从用户接收到查询，并访问多个查询重写路径。所接收的查询与包括在所访问的查询重构路径中的查询进行比较，以识别所接收的查询与包括在相关查询重写路径中的查询之间的关系。基于相关的查询重写路径确定一个或多个普遍的终点，并且向用户呈现至少一个普遍的终点。

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