-
公开(公告)号:WO2017001939A1
公开(公告)日:2017-01-05
申请号:PCT/IB2016/050522
申请日:2016-02-02
申请人: YANDEX EUROPE AG , YANDEX LLC , YANDEX INC.
IPC分类号: G06F17/30
CPC分类号: G06F17/30646 , G06F15/18 , G06F17/18 , G06F17/30389 , G06F17/3053 , G06F17/3064 , G06F17/30867 , G06F17/30926 , G06F17/3097
摘要: There is disclosed a method and a system for generating a search query completion suggestion. The method comprises receiving at least a portion of a search query and determining a first query component therein. A suggested second query component related to the first query component is generated, the search query completion suggestion containing the first query component and the suggested second query component. A list of potentially banned words is accessed to determine if the first query component matches any of the potentially banned words maintained therein. It is then determined if the potentially banned word is associated with a ban marker or an unban marker. A list of ban or unban markers respectively is accessed to determine if the suggested second query component matches any of the ban or unban markers maintained therein, the search query completion suggestion being generated or not generated accordingly.
摘要翻译: 公开了一种用于生成搜索查询完成建议的方法和系统。 该方法包括接收搜索查询的至少一部分并确定其中的第一查询组件。 生成与第一查询组件相关的建议的第二查询组件,搜索查询完成建议包含第一查询组件和建议的第二查询组件。 访问可能禁用的字的列表,以确定第一个查询组件是否匹配其中保留的任何潜在的禁止字。 然后确定潜在禁止的字是否与禁止标记或非标记标记相关联。 分别访问禁用或取消标记的列表,以确定建议的第二查询组件是否匹配其中保留的任何禁用或取消标记,搜索查询完成建议被生成或不相应生成。
-
公开(公告)号:WO2016109307A2
公开(公告)日:2016-07-07
申请号:PCT/US2015/067238
申请日:2015-12-22
CPC分类号: G06F17/278 , G06F17/279 , G06F17/30646 , G06F17/30654 , G06F17/30657 , G06F17/30663 , G06F17/30666 , G06F17/30684 , G06F17/30693 , G10L15/22
摘要: Methods and systems are provided for discriminating ambiguous expressions to enhance user experience. For example, a natural language expression may be received by a speech recognition component. The natural language expression may include at least one of words, terms, and phrases of text. A dialog hypothesis set from the natural language expression may be created by using contextual information. In some cases, the dialog hypothesis set has at least two dialog hypotheses. A plurality of dialog responses may be generated for the dialog hypothesis set. The dialog hypothesis set may be ranked based on an analysis of the plurality of the dialog responses. An action may be performed based on ranking the dialog hypothesis set.
摘要翻译: 提供方法和系统用于区分不明确的表达以增强用户体验。 例如,语音识别组件可以接收自然语言表达。 自然语言表达可以包括文本的单词,术语和短语中的至少一个。 可以通过使用上下文信息来创建从自然语言表达式设置的对话假设。 在某些情况下,对话假设集至少有两个对话假设。 可以为对话假设组生成多个对话响应。 对话假设集合可以基于对多个对话响应的分析进行排名。 可以基于对话假设集合的排序来执行动作。
-
公开(公告)号:WO2015200404A1
公开(公告)日:2015-12-30
申请号:PCT/US2015/037299
申请日:2015-06-24
IPC分类号: G06F17/30
CPC分类号: G06F17/30442 , G06F17/30395 , G06F17/3064 , G06F17/30646 , G06F17/30707 , G06F17/30864 , G06F17/3097
摘要: Architecture that enables the grouping of the same or highly similar intents that are discovered through query reformulation, identifies single intent sessions, and then performs classification of the queries within the single session to determine a change in intent. Queries in a search session that are reformulations of an original query are identified, and the reformulations are distinguished from queries that are issued in a similar sequence to the original query, but cover a completely unrelated intent. When given a user query, a set of accurate and appropriate reformulations are determined, and then used. Additionally, the reformulations can be displayed in accordance with an auto-suggestion technology while the user is still typing, and the reformulations can be displayed when the result screen is displayed as related searches ("Related Searches"). The reformulations can also be used when issuing the query to the search engine.
摘要翻译: 支持通过查询重新设计发现的相同或高度相似意图的分组的体系结构,识别单个意图会话,然后在单个会话中对查询进行分类,以确定意图更改。 识别搜索会话中正在重新构建原始查询的查询,并将重新格式与以与原始查询类似的顺序发布的查询进行区分,但覆盖完全不相关的意图。 当给予用户查询时,确定一组准确和适当的重新制定,然后使用。 此外,当用户仍在打字时,可以根据自动建议技术显示重新配置,并且当结果屏幕显示为相关搜索(“相关搜索”)时,可以显示重新设置。 当向搜索引擎发出查询时也可以使用重新配置。
-
公开(公告)号:WO2014055214A2
公开(公告)日:2014-04-10
申请号:PCT/US2013/059368
申请日:2013-09-12
申请人: AOL INC.
IPC分类号: G06F19/00
CPC分类号: G06F17/30646 , G06F17/30702 , G06F17/30861 , G06F17/30867
摘要: Methods and systems are provided for determining whether a search query with an observed number of occurrences in a set of search queries is a local search query. In accordance with one implementation, a method is provided that comprises determining an expected number of occurrences of a search query and comparing the expected number of occurrences to a threshold. Further, the method includes determining whether the search query is a local search query based, at least in part, on the comparison.
摘要翻译: 提供了用于确定在一组搜索查询中具有观察到的出现次数的搜索查询是否是本地搜索查询的方法和系统。 根据一个实施方式,提供了一种方法,该方法包括确定搜索查询的预期出现次数并将预期出现次数与阈值进行比较。 此外,该方法包括至少部分地基于比较来确定搜索查询是否是本地搜索查询。 p>
-
公开(公告)号:WO2014014724A1
公开(公告)日:2014-01-23
申请号:PCT/US2013/049975
申请日:2013-07-10
发明人: B'FAR, Reza , SPAULDING, Kent , CRANE, Patrick
CPC分类号: G06F21/6263 , G06F17/30395 , G06F17/30646 , G06F21/50 , G06F21/6254 , G06F2221/031 , H04L29/06639 , H04L63/0421
摘要: Techniques for enhancing electronic privacy utilize noise to prevent third parties from determining certain information based on search queries. Users submit search queries as part of their normal activities. For a user, the search queries submitted and information regarding search results used to generate additional search queries on different, but related topics. The generated additional search queries are submitted automatically on behalf of the user at a sufficient frequency to prevent high accuracy data analysis on search queries.
摘要翻译: 用于增强电子隐私的技术利用噪声来防止第三方基于搜索查询确定某些信息。 用户提交搜索查询作为其正常活动的一部分。 对于用户,提交的搜索查询和关于搜索结果的信息用于在不同但相关的主题上生成其他搜索查询。 生成的附加搜索查询以足够的频率代表用户自动提交,以防止对搜索查询的高精度数据分析。
-
公开(公告)号:WO2013088420A4
公开(公告)日:2013-08-08
申请号:PCT/IB2012057360
申请日:2012-12-17
申请人: PYRAMID ANALYTICS BV
发明人: PEREZ AVI , OCHTMAN HERBERT , KOHL OMRI
IPC分类号: G06F17/30
CPC分类号: G06F17/30967 , G06F17/30126 , G06F17/30592 , G06F17/30646 , G06F17/30657
摘要: A user of a computer system is presented an initial presentation of a database query. The query has two or more dimensions, each of which includes two or more elements, and data corresponding to n-tuples of the elements. A selection is received from the user of one or more sets of elements of one of the dimension to transform into (a) (respective) parameter(s). One or more instructions are received from the user to modify the presentation, with each instruction being confined to providing an instance of the active value of a respective parameter. The initial presentation is modified in accordance with only that/those instruction(s).
摘要翻译: 向计算机系统的用户呈现数据库查询的初始呈现。 该查询具有两个或更多个维度,每个维度包括两个或更多元素以及与元素的n元组相对应的数据。 从用户接收维度之一的一个或多个元素集合的选择以变换成(a)(各自的)参数。 从用户接收一个或多个指令以修改表示,其中每个指令被限制为提供相应参数的活动值的实例。 初始演示文稿仅根据那些/那些指令进行修改。
-
公开(公告)号:WO2013067237A3
公开(公告)日:2013-07-04
申请号:PCT/US2012063134
申请日:2012-11-02
申请人: MICROSOFT CORP
IPC分类号: G06F17/30
CPC分类号: G06F17/30696 , G06F17/30566 , G06F17/30646 , G06F17/30654 , G06F17/30672 , G06F17/30864
摘要: Systems and method for routing search query results in a networked computing environment. An initial search query is reformulated into at least one sub-query in accordance with one or more configurable rules. The sub-query is sent to at least one information system or source, and any potential hits associated with the same are optionally combined and then rendered for viewing.
摘要翻译: 在网络计算环境中路由搜索查询结果的系统和方法。 初始搜索查询根据一个或多个可配置规则被重新格式化为至少一个子查询。 子查询被发送到至少一个信息系统或源,并且与之相关联的任何潜在命中可选地组合,然后呈现以供观看。
-
公开(公告)号:WO2013086998A1
公开(公告)日:2013-06-20
申请号:PCT/CN2012/086562
申请日:2012-12-13
申请人: 北大方正集团有限公司 , 北京大学 , 北京北大方正电子有限公司
IPC分类号: G06F17/30
CPC分类号: G06N5/022 , G06F17/278 , G06F17/30604 , G06F17/30646
摘要: 本申请公开了一种用于识别命名实体的识别模型生成方法和装置、以及一种命名实体识别的方法和装置,所述命名实体识别方法包括:获得待训练文本的第一特征信息集;基于第一识别模型对待训练文本的第一特征信息集进行识别,获得第二特征信息集,所述第二特征信息集包含通过所述第一识别模型对所述第一特征信息集进行识别而获得的M个命名实体,所述M为大于或等于零的整数;基于错误驱动模型对所述第二特征信息集中的所述M个命名实体进行错误纠正,获得K个命名实体,所述K为大于或等于零、但小于等于M的整数。
-
公开(公告)号:WO2012016194A1
公开(公告)日:2012-02-02
申请号:PCT/US2011/045980
申请日:2011-07-29
申请人: HASAN, Mohammad Al , PARIKH, Nishith , SINGH, Gyanit , SUNDARESAN, Neelakantan , JOHNSON, Brian S. , KHURANA, Udayan
发明人: HASAN, Mohammad Al , PARIKH, Nishith , SINGH, Gyanit , SUNDARESAN, Neelakantan , JOHNSON, Brian S. , KHURANA, Udayan
IPC分类号: G06F7/00
CPC分类号: G06Q30/0625 , G06F17/30386 , G06F17/3053 , G06F17/3064 , G06F17/30646 , G06F17/30861 , G06Q30/02
摘要: Providing query suggestions using a query log that includes a number of user sessions. The sessions comprise training data including a sequence of a plurality of sets of queries, some of the sets of queries including query transitions followed by a purchase related event. The query log is cleaned and normalized and stationary scores and transition scores of at least some of the plurality of sets is generated. A set of query suggestions is built and similarity scores are computed for at least some of the set of query suggestions to determine whether individual ones of the at least some of the set of query suggestions meet a predetermined assurance level. Those that meet the level are included as elements of the set of query suggestions that meet the assurance level. The set of query suggestions are mixed and ranked in accordance with a user behavior sought to be optimized.
摘要翻译: 使用包含多个用户会话的查询日志提供查询建议。 会话包括包括多组查询的序列的训练数据,其中一些查询集合包括跟随购买相关事件的查询转换。 对查询日志进行清理和归一化,生成多个集合中的至少一些的静态分数和转换分数。 构建一组查询建议,并且针对所述一组查询建议中的至少一些计算相似性分数,以确定所述一组查询建议中的至少一些是否满足预定保证级别。 满足级别的那些包含在满足保证级别的查询建议集合的元素之中。 一组查询建议是混合的,并根据用户行为进行排序,寻求优化。
-
公开(公告)号:WO2010141799A2
公开(公告)日:2010-12-09
申请号:PCT/US2010/037368
申请日:2010-06-04
IPC分类号: G06F17/30
CPC分类号: G06F17/30648 , G06F17/30011 , G06F17/30646 , G06F17/30663 , G06F17/30675 , G06F17/30696 , G06F17/30722 , G06F17/30864 , G06Q10/10 , G06Q50/18
摘要: Systems and techniques are disclosed to rank documents by analyzing a query log generated by a search engine. The query log includes data relating to user behavior, queries and documents. The systems and techniques distill query log information into surrogate documents and extract features from these surrogate documents to rank the documents.
摘要翻译: 公开了通过分析由搜索引擎生成的查询日志来对文档进行排序的系统和技术。 查询日志包括与用户行为,查询和文档相关的数据。 系统和技术将查询日志信息提取为代理文档,并从这些代理文档中提取特征以对文档进行排序。
-
-
-
-
-
-
-
-
-