公开(公告)号：US20240321011A1

公开(公告)日：2024-09-26

申请号：US18631778

申请日：2024-04-10

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Ryo ISHII ,  Ryuichiro Higashinaka ,  Taichi Katayama ,  Junji Tomita ,  Nozomi Kobayashi ,  Kyosuke Nishida

IPC分类号： G06V40/20 ,  G06N20/00 ,  G06V40/10 ,  G10L15/22

CPC分类号： G06V40/28 ,  G06N20/00 ,  G06V40/10 ,  G10L15/22 ,  G10L2015/225

摘要： A nonverbal information generation apparatus includes a nonverbal information generation unit that generates time-information-stamped nonverbal information that corresponds to time-information-stamped text feature quantities on the basis of the time-information-stamped text feature quantities and a learned nonverbal information generation model. The time-information-stamped text feature quantities are configured to include feature quantities that have been extracted from text and time information representing times assigned to predetermined units of the text. The nonverbal information is information for controlling an expression unit that expresses behavior that corresponds to the text.

公开(公告)号：US11651166B2

公开(公告)日：2023-05-16

申请号：US16977422

申请日：2019-02-22

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Itsumi Saito ,  Kyosuke Nishida ,  Hisako Asano ,  Junji Tomita

IPC分类号： G06F40/40 ,  G06F40/20 ,  G06N3/02

CPC分类号： G06F40/40 ,  G06F40/20 ,  G06N3/02

摘要： A learning device of a phrase generation model includes a memory; and a processor configured to execute learning the phrase generation model including an encoder and a decoder, by using, as training data, a 3-tuple. The 3-tuple includes a combination of phrases and at least one of a conjunctive expression representing a relationship between the phrases, and a relational label indicating the relationship represented by the conjunctive expression. The encoder is configured to convert a phrase into a vector from a 2-tuple. The 2-tuple includes a phrase and at least one of the conjunctive expression and the relational label. The decoder is configured to generate, from the converted vector and the conjunctive expression or the relational label, a phrase having the relationship represented by the conjunctive expression or the relational label with respect to the phrase.

公开(公告)号：US12001465B2

公开(公告)日：2024-06-04

申请号：US17050330

申请日：2019-04-10

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Ryuichiro Higashinaka ,  Masahiro Mizukami ,  Junji Tomita

IPC分类号： G06F16/00 ,  G06F16/33 ,  G06F16/332 ,  G06F16/93 ,  G06F40/30

CPC分类号： G06F16/3344 ,  G06F16/3329 ,  G06F16/3334 ,  G06F16/3346 ,  G06F16/93 ,  G06F40/30

摘要： A response selecting apparatus includes a recording part, a document searching part, an information acquiring part, a score calculating part, and a ranking part. The document searching part searches for a question-answer pair from the question-answer pairs recorded in the recording part using the input question as input. The information acquiring part acquires information for quantifying appropriateness of the search-result-question-answer pair with respect to the input question using the input question and the search-result-question-answer pair as input. The score calculating part calculates a score with respect to the input question for each of the search-result-question-answer pairs from a numerical value indicating appropriateness based on a search scheme and a numerical value based on the quantification information. The ranking part selects the search-result-question-answer pairs in descending order of appropriateness indicated by the scores and outputs the answers of the selected question-answer pairs as responses.

公开(公告)号：US11989976B2

公开(公告)日：2024-05-21

申请号：US16969765

申请日：2019-02-15

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Ryo Ishii ,  Ryuichiro Higashinaka ,  Taichi Katayama ,  Junji Tomita ,  Nozomi Kobayashi ,  Kyosuke Nishida

IPC分类号： G06V40/20 ,  G06N20/00 ,  G06V40/10 ,  G10L15/22

CPC分类号： G06V40/28 ,  G06N20/00 ,  G06V40/10 ,  G10L15/22 ,  G10L2015/225

摘要： A nonverbal information generation apparatus includes a nonverbal information generation unit that generates time-information-stamped nonverbal information that corresponds to time-information-stamped text feature quantities on the basis of the time-information-stamped text feature quantities and a learned nonverbal information generation model. The time-information-stamped text feature quantities are configured to include feature quantities that have been extracted from text and time information representing times assigned to predetermined units of the text. The nonverbal information is information for controlling an expression unit that expresses behavior that corresponds to the text.

公开(公告)号：US11537790B2

公开(公告)日：2022-12-27

申请号：US17046668

申请日：2019-04-04

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Katsuji Bessho ,  Hisako Asano ,  Junji Tomita

IPC分类号： G06F17/00 ,  G06F40/237 ,  G06F40/242 ,  G06F40/30

摘要： To arrange all words so that the distance of a given word pair will be appropriate.
Using as input a concept base 22 which is a set of pairs of a word and a vector representing a concept of the word, and a dictionary 24 which is a set of semantically distant or close word pairs, when a word pair C being a pair of given words A, B in the concept base 22 is present in the dictionary 24, conversion means 30 associates with the word pair C a magnitude D of a difference vector between a difference vector V′ between a converted vector of the word A and a converted vector of the word B, and a vector kV determined by multiplying a difference vector V between the vector of the word A in the concept base 22 and the vector of the word B in the concept base 22 by a scalar value k. When the word pair C is not present in the dictionary 24, the conversion means 30 associates the magnitude D of the difference vector between the difference vector V′ and the difference vector V with the word pair C. The conversion means 30 converts the vector of a given word in the concept base 22 such that a total sum of the magnitude D corresponding to every word pair C is minimized.

公开(公告)号：US11036935B2

公开(公告)日：2021-06-15

申请号：US16970334

申请日：2019-02-15

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Ko Mitsuda ,  Ryuichiro Higashinaka ,  Junji Tomita

IPC分类号： G06F17/00 ,  G06F40/186 ,  G06F16/903

摘要： An argumentation structure can be automatically extended.
A template creation unit 22 creates a template using a predetermined rewriting pattern by receiving, as an input, a proposition represented by each of nodes in an argumentation structure including the nodes respectively representing a main proposition of argumentation and a proposition that functions as an argument for another proposition. An additional proposition output unit 28 creates a proposition functioning as an argument for the input proposition using the created template and outputs the created proposition as a proposition represented by a node to be added to the argumentation structure.

公开(公告)号：US12056168B2

公开(公告)日：2024-08-06

申请号：US17795868

申请日：2020-01-29

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Taku Hasegawa ,  Kyosuke Nishida ,  Junji Tomita ,  Hisako Asano

IPC分类号： G06F16/31 ,  G06F16/33 ,  G06N3/02

CPC分类号： G06F16/3331 ,  G06F16/31 ,  G06N3/02

摘要： A learning apparatus according to an embodiment has a feature generation means configured to take a search query, a first document related to the search query, and a second document that is not related to the search query as input, and generate a feature of the search query, a feature of the first document, and a feature of the second document, by using model parameters of a neural network, and an update means configured to take the feature of the search query, the feature of the first document, and the feature of the second document as input, and update the model parameters by using an error function including a cost function that is a differentiable approximation function of an L0 norm.

公开(公告)号：US12026472B2

公开(公告)日：2024-07-02

申请号：US17613417

申请日：2019-05-28

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Yasuhito Osugi ,  Itsumi Saito ,  Kyosuke Nishida ,  Hisako Asano ,  Junji Tomita

IPC分类号： G06F40/35 ,  G06F40/56 ,  G06V30/41

CPC分类号： G06F40/35 ,  G06F40/56 ,  G06V30/41

摘要： A generation unit that takes a question Qi that is a word sequence representing a current question in a dialogue, a document P used to generate an answer Ai to the question Qi, a question history {Qi-1, . . . , Qi-k} that is a set of word sequences representing k past questions, and an answer history {Ai-1, . . . , Ai-k} that is a set of word sequences representing answers to the k questions as inputs, and generates the answer Ai by machine reading comprehension in an extractive mode or a generative mode using pre-trained model parameters is provided.

公开(公告)号：US11972365B2

公开(公告)日：2024-04-30

申请号：US16972187

申请日：2019-04-25

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Atsushi Otsuka ,  Kyosuke Nishida ,  Itsumi Saito ,  Kosuke Nishida ,  Hisako Asano ,  Junji Tomita

IPC分类号： G06F16/2453 ,  G06F16/903 ,  G06N5/04 ,  G06N20/00

CPC分类号： G06N5/04 ,  G06F16/24534 ,  G06F16/90335 ,  G06N20/00

摘要： A question generation device includes: generating means which uses a query and a relevant document including an answer to the query as input and, using a machine learning model having been learned in advance, generates a revised query in which a potentially defective portion of the query is supplemented with a word included in a prescribed lexical set.

公开(公告)号：US11481560B2

公开(公告)日：2022-10-25

申请号：US16767397

申请日：2018-10-05

申请人： NIPPON TELEGRAPH AND TELEPHONE CORPORATION

发明人： Kyosuke Nishida ,  Itsumi Saito ,  Atsushi Otsuka ,  Hisako Asano ,  Junji Tomita

IPC分类号： G06N3/04 ,  G06F40/30 ,  G06F16/93 ,  G06F40/279 ,  G06N3/08

摘要： An information processing device includes a processing unit configured to receive as input a document and a question, and to execute processing to output an answer range as a range of a string that can be an answer to the question in the document, or an answer suitability of the document with respect to the question, by using neural networks, wherein the processing unit includes a first neural network configured to calculate the answer range, and a second neural network configured to calculate the answer suitability, and between the first neural network and the second neural network, part of layers constituting both neural networks is shared.