公开(公告)号：JP2008146283A

公开(公告)日：2008-06-26

申请号：JP2006331474

申请日：2006-12-08

Applicant: Sony Corp ,  ソニー株式会社

Inventor： SAITO MARI ,  YAMAMOTO NORIYUKI

IPC: G06F17/30

CPC classification number: G06Q99/00

Abstract: PROBLEM TO BE SOLVED: To set information expressing the perception of a person to a content as metadata.
SOLUTION: A headgear 2 is a device worn on a head by a test subject, irradiates a head of a person to be tested with near-infrared light, and measures a hemoglobin amount responding to oxygen consumption by the activity of a brain of the person to be tested listening to a music content as a biological response. The setting of the metadata by this information processor 1 is performed by predicting the biological response thought that the person to be tested probably shows if making the person to be tested actually listen based on bioinformation obtained by making a plurality of the persons to be tested actually listen to the music contents of a limited number such as 100 pieces of music, and setting information expressing the predicted biological response to the target music content as the metadata. The information processor can be applied to a device setting metadata to a content.
COPYRIGHT: (C)2008,JPO&INPIT

Abstract translation: 要解决的问题：将表示人的感知的信息设置为内容作为元数据。 头巾2是由被检对象佩戴在头部的装置，用近红外光照射被测试者的头部，通过脑的活动测定与氧消耗相应的血红蛋白量 的被测试者听音乐内容作为生物反应。 该信息处理器1的元数据的设置是通过预测生物反应的思想来进行的，该生物学反应认为被测试者可能基于通过使多个被测试者实际获得的生物信息实际上听从而进行测试 收听有限数量的音乐内容，例如100首音乐，并将表示预测的生物学反应的信息设置为目标音乐内容作为元数据。 信息处理器可以应用于将元数据设置为内容的设备。 版权所有（C）2008，JPO＆INPIT

公开(公告)号：JP2007202181A

公开(公告)日：2007-08-09

申请号：JP2007054676

申请日：2007-03-05

Applicant: Sony Corp ,  ソニー株式会社

Inventor： YAMAMOTO NORIYUKI ,  MIYAZAKI MITSUHIRO ,  SAITO MARI ,  KOIKE HIROYUKI

IPC: G06F17/30 ,  H04N7/173 ,  H04N21/235

CPC classification number: H04N21/2353 ,  H04N21/435 ,  H04N21/44213 ,  H04N21/4668 ,  H04N21/84

Abstract: PROBLEM TO BE SOLVED: To generate a program vector representing an attribute of a program applicable to a distribution server which distributes contents. SOLUTION: In step S11, EPG data are provided and in step S12, meta data required for generating a program vector PP are extracted. In step S13, morpheme analysis is applied to a title, contents and the like contained in the meta data to decompose them into words and in step S14, items contained in the meta data are vectorized to generate the program vector PP. In step S15, a corresponding effect vector is extracted based on a category of the program of the provided meta data. In step S16, the extracted effect vector and the generated vector are associated with each other, and processing is finished. COPYRIGHT: (C)2007,JPO&INPIT

Abstract translation: 要解决的问题：生成表示可应用于分发内容的分发服务器的程序的属性的程序向量。 解决方案：在步骤S11中，提供EPG数据，并且在步骤S12中，提取用于生成节目向量PP所需的元数据。 在步骤S13中，对包含在元数据中的标题，内容等应用语素分析，将其分解成单词，在步骤S14中，包含在元数据中的项目被矢量化，生成程序向量PP。 在步骤S15中，基于提供的元数据的程序的类别来提取相应的效果向量。 在步骤S16中，提取的效果矢量和生成的矢量相互关联，处理结束。 版权所有（C）2007，JPO＆INPIT

公开(公告)号：JP2006300970A

公开(公告)日：2006-11-02

申请号：JP2005117710

申请日：2005-04-15

Applicant: Sony Corp ,  ソニー株式会社

Inventor： YAMAMOTO NORIYUKI

IPC: G10K15/02

CPC classification number: G06F17/30761 ,  G06F17/30017 ,  G06F17/30743 ,  G06F17/30772

Abstract: PROBLEM TO BE SOLVED: To generate a play list in which similarities between contents to be reproduced and user's preferences change with specific fluctuations.
SOLUTION: When musical pieces to be reproduced are determined while standards are changed according to a standard table in Fig. 3, a starting musical piece (musical piece at the head in the play list) is determined, for example, by a user's music choice in this example; and a second musical piece to be reproduced right after the starting musical piece is a musical piece whose similarity to a user preference vector is 0.8 to 1.0 (within a similarity range based upon a standard with a standard number 1), a third musical piece is a musical piece whose similarity to the user preference vector is 0.8 to 1.0 (within a similarity range based upon a standard with a standard number 2), and a fourth musical piece is a musical piece whose similarity to the user preference vector is 0.6 to 0.8 (within a similarity range based upon a standard with a standard number 3).
COPYRIGHT: (C)2007,JPO&INPIT

Abstract translation: 要解决的问题：生成播放列表，其中要再现的内容和用户偏好与特定波动之间的相似性变化。

解决方案：当标准根据图1的标准表改变时，确定要再现的乐曲。 如图3所示，例如在本示例中，通过用户的音乐选择确定起始乐曲（播放列表中头部的乐曲） 并且在起始乐曲之后立即再现的第二乐曲是与用户偏好向量的相似度为0.8至1.0（在基于标准编号1的标准的相似范围内）的乐曲，第三乐曲是 与用户偏好向量的相似度为0.8〜1.0的乐曲（在基于标准2的标准的相似范围内），第四乐曲是与用户偏好向量的相似度为0.6〜0.8的乐曲 （在基于标准编号3的标准的相似范围内）。 版权所有（C）2007，JPO＆INPIT

公开(公告)号：JP2006293246A

公开(公告)日：2006-10-26

申请号：JP2005117709

申请日：2005-04-15

Applicant: Sony Corp ,  ソニー株式会社

Inventor： YAMAMOTO NORIYUKI

IPC: G10L11/00 ,  G06F17/30 ,  G11B27/10

CPC classification number: G06F17/30761 ,  G06F17/30017 ,  G06F17/30743 ,  G06F17/30772

Abstract: PROBLEM TO BE SOLVED: To generate a play list such that similarities between two successively reproduced contents vary having certain fluctuations.
SOLUTION: Musical pieces to be reproduced are determined while reference is varied on the basis of a reference table. A starting musical piece (musical piece at the head of the play list) is selected by, for example, a user, a 2nd musical piece to be played right after the starting musical piece is a musical piece having a similarity of 0.8 to 1.0 (within a similarity range of reference with a reference number 1) to the starting musical piece, a 3rd musical piece is a musical piece having a similarity of 0.8 to 1.0 (similarity range of reference with a reference number 2) to the 2nd musical piece, and a 4th musical piece is a musical piece having a similarity of 0.6 to 0.8 (similarity range of reference with a reference number 3) to the 3rd musical piece.
COPYRIGHT: (C)2007,JPO&INPIT

Abstract translation: 要解决的问题：产生播放列表，使得两个连续再现的内容之间的相似性变化，具有一定的波动。

解决方案：基于参考表改变参考，确定要再现的乐曲。 起始乐曲（播放列表的头部的乐曲）由例如用户选择，在开始乐曲之后要播放的第二乐曲是相似度为0.8到1.0的乐曲（ 在与参考编号1相关的相似范围内，对于起始乐曲，第3曲是与第2乐曲相似度为0.8〜1.0（与参考编号2相关的相似度范围）的乐曲， 第4曲是与第3乐曲相似度为0.6〜0.8（与参考编号3相关的相似度范围）的乐曲。 版权所有（C）2007，JPO＆INPIT

公开(公告)号：JP2006215643A

公开(公告)日：2006-08-17

申请号：JP2005025298

申请日：2005-02-01

Applicant: Sony Corp ,  ソニー株式会社

Inventor： YAMAMOTO NORIYUKI

IPC: G06F17/30

CPC classification number: G06F17/30029 ,  G06F17/30749 ,  G06F17/30758 ,  Y10S707/99948

Abstract: PROBLEM TO BE SOLVED: To recommend appropriate content that matches the user's selection of content whose attributes are different from those of the content that a user uses usually. SOLUTION: When informed by a control part 12 about a selected song, a short-term preference vector creating part 22 obtains the song vector V of the song from a song vector storage part 21 and stores the song vector V in a short-term preference vector storage part 23 as a short-term preference vector Vnp. A recommending part 26 calculates the difference between the short-term preference vector and a long-term preference vector, determines a song to be recommended to the user by means of weighting based on the difference, and recommends the song to the user. COPYRIGHT: (C)2006,JPO&NCIPI

Abstract translation: 要解决的问题：推荐与用户对属性与用户通常使用的内容不同的内容选择相匹配的适当内容。 解决方案：当由控制部分12通知所选歌曲时，短期偏好向量创建部分22从歌曲向量存储部分21获得歌曲的歌曲矢量V，并将歌曲矢量V存储在短片中 - 优选向量存储部分23作为短期偏好向量Vnp。 推荐部分26计算短期偏好向量和长期偏好向量之间的差异，通过基于差异的加权来确定要推荐给用户的歌曲，并向用户推荐该歌曲。 版权所有（C）2006，JPO＆NCIPI

公开(公告)号：JP2006048286A

公开(公告)日：2006-02-16

申请号：JP2004226788

申请日：2004-08-03

Applicant: Sony Corp ,  ソニー株式会社

Inventor： TATENO HIROSHI ,  YAMAMOTO NORIYUKI ,  SAITO MARI ,  MIYAZAKI MITSUHIRO

IPC: G06F17/30

CPC classification number: G06F17/30997

Abstract: PROBLEM TO BE SOLVED: To make it possible to perform a processing using meta-data such as content recommendation while taking the cooccurrence relationship of meta-date into consideration. SOLUTION: A matrix generation part 18 generates a meta-data matrix consisting of N (N: an integer of ≥1) rows corresponding to N meta-data and M (M: an integer of ≥1) columns corresponding to M contents. An LSA arithmetic part 20 subjects the meta-data matrix to singular value decomposition processing to generate an approximate matrix of the meta-data matrix. A meta-data extraction part 21 calculates index values indicating importance of corresponding meta-data, e.g. characteristic differences for the N meta-data based upon differences between the meta-data matrix and approximate matrix and extracts important meta-data or unnecessary meta-data from the N meta-data in a meta-data storage part 16 according to those index values. This invention is applicable to an information processor which recommends contents. COPYRIGHT: (C)2006,JPO&NCIPI

Abstract translation: 要解决的问题：考虑到元日期的同现关系，使得可以使用诸如内容推荐的元数据来执行处理。 解决方案：矩阵生成部分18生成由对应于N个元数据的N（N：整数≥1）行和M（M：≥1的整数）列组成的元数据矩阵 内容。 LSA算术部分20对元数据矩阵进行奇异值分解处理，以生成元数据矩阵的近似矩阵。 元数据提取部21计算表示对应的元数据的重要性的索引值，例如， 根据元数据矩阵和近似矩阵之间的差异对N个元数据的特征差异，并根据这些索引值从元数据存储部分16中的N个元数据中提取重要的元数据或不必要的元数据 。 本发明适用于推荐内容的信息处理器。 版权所有（C）2006，JPO＆NCIPI

公开(公告)号：JP2005165632A

公开(公告)日：2005-06-23

申请号：JP2003403146

申请日：2003-12-02

Applicant: Sony Corp ,  ソニー株式会社

Inventor： YAMAMOTO NORIYUKI

IPC: G06F7/00 ,  G06F17/00 ,  G06F17/30 ,  G06K9/62

CPC classification number: G06F17/30017 ,  Y10S707/99937

Abstract: PROBLEM TO BE SOLVED: To use the proper taste information of a user. SOLUTION: A matrix calculating part 22 acquires M(M is an integral value which is 1 or more) pieces of feature vectors CCV with the weight values of N(N is an integral value which is 2 or more) pieces of contents meta information or context information as elements, and generates a matrix CCM with the M pieces of feature vectors CCV as column components, and generates an approximate matrix CCM * by updating the weight values of the respective elements of the M pieces of feature vectors CCV in order to reinforce relations between the respective elements of each of the M pieces of feature vectors CCV. A user preference vector(UPV) generating part 23 generates user preference vectors UPV * based on the approximate CCM * . A matching part 3 calculates the similarity of the feature vectors CCV generated from new contents meta information or new context information to the user preference vectors UPV * . This invention may be applied to an information processor which recommends contents. COPYRIGHT: (C)2005,JPO&NCIPI

Abstract translation: 要解决的问题：使用用户的正确口味信息。 解决方案：矩阵计算部分22获得具有权重值N（N是2或更大的整数值）内容的M（M是1或更多个）的特征矢量CCV的M 元信息或上下文信息作为元素，并且生成具有M个特征向量CCV作为列分量的矩阵CCM，并且通过更新相关元素的权重值来生成近似矩阵CCM * M个特征向量CCV，以便加强每个M个特征向量CCV的各个元素之间的关系。 用户偏好向量（UPV）生成部23基于近似CCM * 生成用户偏好向量UPV * 。 匹配部分3计算从新内容元信息或新上下文信息生成的特征向量CCV与用户偏好向量UPV * 的相似度。 本发明可以应用于推荐内容的信息处理器。 版权所有（C）2005，JPO＆NCIPI

公开(公告)号：JP2005159770A

公开(公告)日：2005-06-16

申请号：JP2003396224

申请日：2003-11-26

Applicant: Sony Corp ,  ソニー株式会社

Inventor： FUJIWARA NOBUYUKI ,  NARAHARA TATSUYA ,  YAMAMOTO NORIYUKI ,  SAITO MARI ,  MIYAZAKI MITSUHIRO ,  KOIKE HIROYUKI

IPC: G06F17/00 ,  G06F17/30 ,  G11B31/00 ,  H04N5/44 ,  H04N5/76 ,  H04N5/765 ,  H04N5/77 ,  H04N5/781 ,  H04N5/782 ,  H04N5/91 ,  H04N7/173 ,  H04N9/804

CPC classification number: H04N5/44543 ,  H04N5/765 ,  H04N5/77 ,  H04N5/781 ,  H04N5/782 ,  H04N9/8042 ,  H04N21/4147 ,  H04N21/4334 ,  H04N21/4335 ,  H04N21/435 ,  H04N21/4622 ,  H04N21/482 ,  H04N21/84 ,  H04N21/8455

Abstract: PROBLEM TO BE SOLVED: To allow a user to easily determine programs to be recorded or reserved for recording, with display of information regarding scheduled broadcast programs such as EPG.
SOLUTION: A trailer is recorded or reserved for recording by discriminating metadata of a received broadcast program. If a trailer, broadcasted last time, is held as a recorded data in the storage area of equipment, a user can reproduce it when the user sees EPG of a program for viewing or reserving recording of it. By this reproduction, the user can intuitively judge whether a program which is scheduled to be broadcasted should be viewed or reserved for recording.
COPYRIGHT: (C)2005,JPO&NCIPI

Abstract translation: 要解决的问题：允许用户容易地确定要被记录或预留用于记录的节目，并显示关于诸如EPG的预定广播节目的信息。 解决方案：通过识别所接收的广播节目的元数据来记录或保留预告用于记录。 如果上次广播的预告片作为记录数据被保存在设备的存储区域中，则当用户看到用于查看或保留记录的节目的EPG时，用户可以再现它。 通过该再现，用户可以直观地判断是否应该查看或保留用于记录的被调度广播的节目。 版权所有（C）2005，JPO＆NCIPI

公开(公告)号：JP2005039748A

公开(公告)日：2005-02-10

申请号：JP2003277182

申请日：2003-07-18

Applicant: Sony Corp ,  ソニー株式会社

Inventor： KOIKE HIROYUKI ,  YAMAMOTO NORIYUKI ,  SAITO MARI ,  MIYAZAKI MITSUHIRO ,  NARAHARA TATSUYA ,  FUJIWARA NOBUYUKI

IPC: G06F17/30 ,  H04N5/44 ,  H04N5/91

Abstract: PROBLEM TO BE SOLVED: To recommend a program corresponding to the taste of a user. SOLUTION: The time from an EPG program list display start time to a recording program selection time of a program A is a domain (time (t 1 -t 0 )) illustrated by VS A and similarly, in programs B, C, the times become domains (time (t 2 -t 0 )), (time (t 3 -t 0 )) illustrated by VS B , VS C . At such a time, when the ratio of the lengths of the times is VS A :VS B :VS C =1:4:2 and when X 1 is defined as 4, with respect to weights of categories of the programs A, B, C, "variety" that is the category of the program A is 4, "drama" that is the category of the program B is 1, and "news" that is the category of the program C is 2. The present invention can be applied to a video recorder. COPYRIGHT: (C)2005,JPO&NCIPI

Abstract translation: 要解决的问题：推荐与用户的口味对应的程序。 解决方案：从EPG节目列表显示开始时间到节目A的记录节目选择时间的时间是一个域（时间（t 1 -t 0 ）），并且类似地，在程序B，C中，时间变为域（时间（t SB> 2 -t 0 ）） ，（VS（SB SB））所示的时间（t（SB）3 -t 0 ））。 在这样的时间，当时间长度的比例为VS B 时：VS SB = 1：4：2 并且当X 1 被定义为4时，对于节目A，B，C的类别的权重，作为节目A的类别的“品种”是4，“戏剧”是 程序B的类别为1，作为程序C的类别的“新闻”为2.本发明可以应用于录像机。 版权所有（C）2005，JPO＆NCIPI

公开(公告)号：JP2004355340A

公开(公告)日：2004-12-16

申请号：JP2003152447

申请日：2003-05-29

Applicant: Sony Corp ,  ソニー株式会社

Inventor： SAITO MARI ,  YAMAMOTO NORIYUKI ,  MIYAZAKI MITSUHIRO ,  FUJIWARA NOBUYUKI ,  NARAHARA TATSUYA ,  KOIKE HIROYUKI

IPC: G06F17/30 ,  H04N5/44 ,  H04N5/445 ,  H04N5/76 ,  H04N5/91 ,  H04N7/173 ,  H04N17/00

Abstract: PROBLEM TO BE SOLVED: To easily understandably exhibit a reason why a program is recommended, so as to recommend the program met to a preference of a user.
SOLUTION: A metadata of the program is acquired by a preference information extracting part 101, and preference information of the user is generated based on an attribute of the metadata to be stored in a server. The metadata to be recommended is retrieved by a recommendation information retrieval part 102, based on the preference information of the user, and a recommendation information exhibiting part 103 exhibits to the user the attribute constituting the preference information as the reason of recommendation. The user books picture recording of the program based on the the exhibited recommendation reason. The present invention is applicable for an HDD recorder or a personal computer.
COPYRIGHT: (C)2005,JPO&NCIPI