利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

102 条记录 (耗时 0.037 秒)

    DOCUMENT REPRESENTATION TRANSITIONING
    21.
    发明申请
    DOCUMENT REPRESENTATION TRANSITIONING 有权
    文件陈述过渡

    公开(公告)号：US20110314372A1

    公开(公告)日：2011-12-22

    申请号：US12820297

    申请日：2010-06-22

    申请人： Jaime Teevan ,  Susan T. Dumais ,  Daniel J. Liebling

    发明人： Jaime Teevan ,  Susan T. Dumais ,  Daniel J. Liebling

    IPC分类号： G06F17/30 ,  G06F17/21

    CPC分类号： G06F17/2211 ,  G06F17/211

    摘要： One or more techniques and/or systems are provided for transitioning between representations of an electronic document. Elements, such as visual elements, common between a first set of elements from a first representation of the document and a second set of elements from a second representation of the document are identified. The non-intersecting elements from the first and second sets are respectively ranked in accordance with a representation relevance. First set non-intersecting elements are removed from an intermediate representation of the document, and second set non-intersecting elements are added to the intermediate representation, while the intermediate representation is not equivalent to the second representation; and respective iterations of the intermediate representation are output, such as to a display to depict a transition from the first representation of the document to the second representation of the document.

    摘要翻译： 提供一个或多个技术和/或系统用于在电子文档的表示之间进行转换。 识别来自文档的第一表示的第一组元素与来自文档的第二表示的第二组元素之间的元素，诸如视觉元素。 来自第一和第二组的不相交的元素分别根据表示相关性排列。 首先设置的非相交元素从文档的中间表示中移除，并且将第二组非相交元素添加到中间表示，而中间表示不等同于第二表示; 并且输出中间表示的相应迭代，诸如显示以描绘从文档的第一表示到文档的第二表示的转换。

    Preference judgements for relevance
    22.
    发明授权
    Preference judgements for relevance 有权
    相关性的偏好判断

    公开(公告)号：US08069179B2

    公开(公告)日：2011-11-29

    申请号：US12108531

    申请日：2008-04-24

    申请人： David M. Chickering ,  Paul N. Bennett ,  Susan T. Dumais ,  Benjamin Austin Carterette

    发明人： David M. Chickering ,  Paul N. Bennett ,  Susan T. Dumais ,  Benjamin Austin Carterette

    IPC分类号： G06F7/00 ,  G06F17/30

    CPC分类号： G06F17/30867

    摘要： The claimed subject matter provides a system that trains or evaluates ranking techniques by employing or obtaining relative preference judgments. The system can include mechanisms that retrieve a set of documents from a storage device, combine the set of documents with a query or judgment task received via an interface to form a comparative selection panel, and present the comparative selection panel for evaluation by an assessor. The system further requests the assessor to make a selection as to which document included in the set of documents and presented in the comparative selection panel most satisfies the query or judgment task, and thereafter produces a comparative assessment of the set of documents based on the selections elicited from the assessor and associated with the set of documents.

    摘要翻译： 所要求保护的主题提供了通过采用或获得相对偏好判断来训练或评估排名技术的系统。 该系统可以包括从存储装置检索一组文档的机构，将该组文件与通过界面接收到的查询或判断任务组合以形成比较选择面板，并呈现评估者进行评估的比较选择面板。 该系统进一步要求评估人员选择一组文件中包含的文件，并在比较选择面板中提供的文档最符合查询或判断任务，然后根据选择对文档集进行比较评估 从评估员处获得并与该组文件相关联。

    Cost-benefit approach to automatically composing answers to questions by extracting information from large unstructured corpora
    24.
    发明授权
    Cost-benefit approach to automatically composing answers to questions by extracting information from large unstructured corpora 失效
    通过从大型非结构化语料库中提取信息来自动构成问题答案的成本效益方法

    公开(公告)号：US07739215B2

    公开(公告)日：2010-06-15

    申请号：US12417959

    申请日：2009-04-03

    申请人： Eric J. Horvitz ,  David R. Azari ,  Susan T. Dumais ,  Eric D. Brill

    发明人： Eric J. Horvitz ,  David R. Azari ,  Susan T. Dumais ,  Eric D. Brill

    IPC分类号： G06F17/00 ,  G06N5/02

    CPC分类号： G06F17/30684 ,  G06F17/30687 ,  Y10S707/99933

    摘要： The present invention relates to a system and methodology to facilitate extraction of information from a large unstructured corpora such as from the World Wide Web and/or other unstructured sources. Information in the form of answers to questions can be automatically composed from such sources via probabilistic models and cost-benefit analyses to guide resource-intensive information-extraction procedures employed by a knowledge-based question answering system. The analyses can leverage predictions of the ultimate quality of answers generated by the system provided by Bayesian or other statistical models. Such predictions, when coupled with a utility model can provide the system with the ability to make decisions about the number of queries issued to a search engine (or engines), given the cost of queries and the expected value of query results in refining an ultimate answer. Given a preference model, information extraction actions can be taken with the highest expected utility. In this manner, the accuracy of answers to questions can be balanced with the cost of information extraction and analysis to compose the answers.

    摘要翻译： 本发明涉及一种便利从诸如万维网和/或其他非结构化来源的大型非结构化语料库提取信息的系统和方法。 通过概率模型和成本效益分析，可以通过这些来源自动构成问题答案形式的信息，以指导基于知识的问答系统采用的资源密集型信息提取程序。 分析可以利用由贝叶斯或其他统计模型提供的系统生成的答案的最终质量的预测。 当与实用新型相结合时，这种预测可以为系统提供对发出给搜索引擎（或引擎）的查询数量的决定的能力，考虑到查询的成本和查询结果的期望值来提炼最终的 回答。 给定一个偏好模型，可以采用最高预期效用的信息提取动作。 以这种方式，可以将问题答案的准确性与信息提取和分析的成本进行平衡，以构成答案。

    Activity-ware for non-textual objects
    25.
    发明授权
    Activity-ware for non-textual objects 有权
    非文本对象的活动零件

    公开(公告)号：US07716054B2

    公开(公告)日：2010-05-11

    申请号：US11771135

    申请日：2007-06-29

    申请人： Jensen M. Harris ,  Susan T. Dumais ,  Richard J. Wolf ,  Eran Megiddo

    发明人： Jensen M. Harris ,  Susan T. Dumais ,  Richard J. Wolf ,  Eran Megiddo

    IPC分类号： G10L21/00

    CPC分类号： G06F17/30758 ,  G06F17/30743 ,  G06F17/30749 ,  G10L17/26 ,  G10L25/48 ,  G11B27/28 ,  G11B27/329

    摘要： Providing for summarization and analysis of audio content is described herein. By way of example, an oral conversation can be analyzed, such that points of interest within the oral conversation can be identified and file locations related to such points of interest can be marked. Points of interest can be inferred based on a level of energy, e.g., excitement, pitch, tone, pace, or the like, associated with one or more speakers. Alternatively, or in addition, speaker and/or reviewer activity can form the basis for identifying points of interest within the conversation. Moreover, a compilation of the identified points of interest and portions of the original oral conversation related thereto can be assembled. As described herein, audio content can be succinctly summarized with respect to inferred and/or indicated points of interest, to facilitate an efficient and pertinent review of such content.

    摘要翻译： 本文描述了对音频内容的总结和分析的提供。 作为示例，可以分析口头对话，使得可以识别口头对话内的兴趣点，并且可以标记与这些兴趣点相关的文件位置。 可以基于与一个或多个扬声器相关联的能量水平，例如兴奋，音调，音调，步调等来推断兴趣点。 或者或另外，说话者和/或审阅者活动可以形成用于识别会话内的兴趣点的基础。 此外，可以汇编所识别的兴趣点和与之相关的原始口头对话的部分。 如本文所述，音频内容可以相对于推断的和/或指示的兴趣点被简明地概括，以便于对这些内容的有效和有针对性的审查。

    COMMUNICATION WORKSPACE
    26.
    发明申请
    COMMUNICATION WORKSPACE 有权
    通信工作空间

    公开(公告)号：US20090254390A1

    公开(公告)日：2009-10-08

    申请号：US12098027

    申请日：2008-04-04

    申请人： Eran Megiddo ,  Joshua T. Goodman ,  Jensen M. Harris ,  Susan T. Dumais ,  Michael V. Ehrenberg ,  Srikanth Shoroff ,  William H. Gates, III

    发明人： Eran Megiddo ,  Joshua T. Goodman ,  Jensen M. Harris ,  Susan T. Dumais ,  Michael V. Ehrenberg ,  Srikanth Shoroff ,  William H. Gates, III

    IPC分类号： G06Q99/00 ,  G06Q10/00 ,  G06F21/00

    CPC分类号： G06Q10/10 ,  G06F21/6218

    摘要： Multiple pieces of information can be arranged into a single construct that allows the employee to ascertain information quickly while at her workstation. Selection of information for placement into the construct can employ various statistical models and the like. Selective pieces of information can be masked for a user's construct based upon access rights of the user. Constructs can be configured by a user based on personal preferences as well as by an administrator. Population of metadata upon the construct can be performed automatically through an instruction of the administrator or be overridden by a user request. In addition, various types of synchronization can be implemented between constructs, such that identical or near-identical information is populated upon multiple constructs.

    摘要翻译： 多个信息可以被安排成一个单一的结构，允许员工在她的工作站快速地确定信息。 用于放置到构造中的信息的选择可以采用各种统计模型等。 基于用户的访问权限，可以为用户的构造掩蔽选择性的信息。 构造可以由用户基于个人偏好以及由管理员进行配置。 可以通过管理员的指令自动执行构造中的元数据的群体，或者被用户请求覆盖。 此外，可以在构造之间实现各种类型的同步，使得在多个构造上填充相同或近乎相同的信息。

    Cost-benefit approach to automatically composing answers to questions by extracting information from large unstructured corpora
    27.
    发明授权
    Cost-benefit approach to automatically composing answers to questions by extracting information from large unstructured corpora 有权
    通过从大型非结构化语料库中提取信息来自动构成问题答案的成本效益方法

    公开(公告)号：US07516113B2

    公开(公告)日：2009-04-07

    申请号：US11469136

    申请日：2006-08-31

    申请人： Eric J. Horvitz ,  David R. Azari ,  Susan T. Dumais ,  Eric D. Brill

    发明人： Eric J. Horvitz ,  David R. Azari ,  Susan T. Dumais ,  Eric D. Brill

    IPC分类号： G06F17/00 ,  G06N5/02

    CPC分类号： G06F17/30684 ,  G06F17/30687 ,  Y10S707/99933

    摘要： The present invention relates to a system and methodology to facilitate extraction of information from a large unstructured corpora such as from the World Wide Web and/or other unstructured sources. Information in the form of answers to questions can be automatically composed from such sources via probabilistic models and cost-benefit analyses to guide resource-intensive information-extraction procedures employed by a knowledge-based question answering system. The analyses can leverage predictions of the ultimate quality of answers generated by the system provided by Bayesian or other statistical models. Such predictions, when coupled with a utility model can provide the system with the ability to make decisions about the number of queries issued to a search engine (or engines), given the cost of queries and the expected value of query results in refining an ultimate answer. Given a preference model, information extraction actions can be taken with the highest expected utility. In this manner, the accuracy of answers to questions can be balanced with the cost of information extraction and analysis to compose the answers.

    摘要翻译： 本发明涉及一种便利从诸如万维网和/或其他非结构化来源的大型非结构化语料库提取信息的系统和方法。 通过概率模型和成本效益分析，可以通过这些来源自动构成问题答案形式的信息，以指导基于知识的问答系统采用的资源密集型信息提取程序。 分析可以利用由贝叶斯或其他统计模型提供的系统生成的答案的最终质量的预测。 当与实用新型相结合时，这种预测可以为系统提供对发出到搜索引擎（或引擎）的查询数量的决定的能力，考虑到查询的成本和查询结果的期望值来提炼最终的 回答。 给定一个偏好模型，可以采用最高预期效用的信息提取动作。 以这种方式，可以将问题答案的准确性与信息提取和分析的成本进行平衡，以构成答案。