Information processing method, information processing device, and program

Abstract

To extract content that evokes a user's prescribed feeling with high accuracy.SOLUTION: An information processing method makes a processor included in an information processing device: sequentially acquire brain wave signals of a prescribed user stimulated by prescribed content from a brain wave measurement device attached to the prescribed user; input the brain wave signals that are sequentially acquired to a brain wave model for inputting the brain wave signals of the prescribed user stimulated by the content to predict an evocation degree to evoke a prescribed feeling of the prescribed user by the content in stimulus and predict the evocation degree of the prescribed feeling by the prescribed content; input a feature amount of the content with a correspondence relation between the feature amount of the prescribed content and the evocation degree predicted by the brain wave model as relearning data to update the content model for predicting an evocation degree; and extract the prescribed content that evokes the prescribed feeling of the prescribed user on the basis of the evocation degree predicted by the content model from in a content list.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an information processing method, an information processing device, and a program. [Background technology]

[0002] Conventionally, there is a known technology that can control the user's emotions and make them listen to enjoyable music by estimating the user's emotions from brain wave signals and playing music that matches those emotions ( For example, see Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0003] [Non-patent document 1] Ehrlich SK, Agres KR, Guan C, Cheng G (2019), "A closed-loop, music-based brain-computer interface for emotion mediation", [online], March 18, 2019, PLOS ONE, [Reiwa 5 Retrieved June 30], Internet <URL: https: / / doi.org / 10.1371 / journal.pone.0213516> [Summary of the invention] [Problem to be solved by the invention]

[0004] Since it depends on the individual what kind of music has the effect of regulating emotions, conventional technology is capable of adjusting the user's predetermined emotions by playing appropriate music according to the individual user. It became.

[0005] Therefore, one aspect of the disclosed technology aims to provide an information processing method, an information processing device, and a program that can extract content that evokes a predetermined emotion in a user with high precision. [Means to solve the problem]

[0006] An information processing method according to one aspect of the disclosed technology includes: a processor included in an information processing apparatus sequentially acquiring brain wave signals of a predetermined user stimulated by predetermined content from an electroencephalogram measurement device worn by the predetermined user; The acquired brain wave signals are sequentially input into a brain wave model that predicts the degree of arousal of a predetermined emotion in the predetermined user by the content being stimulated by inputting the brain wave signals of a predetermined user stimulated by the content. Predicting the degree of arousal of a predetermined emotion by content, and predicting the degree of arousal by inputting the feature amount of the content using the correspondence between the feature amount of the predetermined content and the degree of arousal predicted by the electroencephalogram model as relearning data. and extracting from the content list a predetermined content that evokes a predetermined emotion in a predetermined user based on the degree of arousal predicted by the content model.

[0007] An information processing apparatus including a processor according to an aspect of the disclosed technology includes the following steps: the processor sequentially acquires brain wave signals of a predetermined user stimulated by predetermined content from an electroencephalogram measuring device worn by the predetermined user; The acquired brain wave signals are sequentially input to an brain wave model that predicts the degree of arousal of a predetermined emotion in a predetermined user by the content being stimulated. Predicting the degree of arousal of a predetermined emotion by content, and predicting the degree of arousal by inputting the feature amount of the content using the correspondence between the feature amount of the predetermined content and the degree of arousal predicted by the electroencephalogram model as relearning data. and extracting from the content list a predetermined content that arouses a predetermined emotion in a predetermined user based on the degree of arousal predicted by the content model.

[0008] A program according to an aspect of the disclosed technology causes a processor included in an information processing apparatus to sequentially acquire brain wave signals of a predetermined user stimulated by predetermined content from an electroencephalogram measurement device worn by the predetermined user; The acquired brain wave signals are sequentially input to an brain wave model that predicts the degree of arousal of a predetermined emotion in the predetermined user by the content being stimulated by inputting the brain wave signals of the predetermined user to be stimulated. Predicting the degree of arousal of a predetermined emotion by content, and predicting the degree of arousal by inputting the feature values ​​of the content, using the correspondence between the feature values ​​of the predetermined content and the degree of arousal predicted by the electroencephalogram model as relearning data. and extracting from the content list a predetermined content that evokes a predetermined emotion in a predetermined user based on the degree of arousal predicted by the content model. [Effect of the invention]

[0009] According to the present invention, it is possible to provide an information processing method, an information processing device, and a program that can extract content that evokes a predetermined emotion in a user with high precision. [Brief explanation of the drawing]

[0010] [Figure 1] 1 is a diagram illustrating an example outline of an information processing system 1 according to an embodiment. [Figure 2] FIG. 1 is a diagram showing an example of an earphone set 10 according to an embodiment. [Figure 3] FIG. 2 is a diagram illustrating an example of a schematic cross section of an earphone 100R according to an embodiment. [Figure 4] FIG. 3 is a block diagram showing an example of an information processing device 30 according to an embodiment. [Figure 5A] FIG. 2 is a diagram schematically showing a content model. [Figure 5B] FIG. 2 is a diagram schematically showing an electroencephalogram model. [Figure 6] FIG. 2 is a block diagram illustrating an example of an information processing device 50 according to an embodiment. [Figure 7] 3 is a flowchart illustrating an example of processing in the information processing system 1. FIG. [Figure 8] 3 is a flowchart illustrating an example of processing in the information processing system 1. FIG. [Figure 9] 5 is a flowchart showing an example of processing in the information processing system 1. FIG. [Figure 10] 3 is a flowchart illustrating an example of processing in the information processing system 1. FIG. [Details for carrying out the invention]

[0011] Embodiments of the present invention will be described below with reference to the drawings. However, the embodiments described below are merely examples, and there is no intention to exclude the application of various modifications and techniques not specified below. That is, the present invention can be implemented with various modifications without departing from the spirit thereof. In addition, in the description of the drawings below, the same or similar parts are denoted by the same or similar symbols. The drawings are schematic and do not necessarily correspond to actual dimensions or proportions. The drawings may also include portions that differ in their dimensional relationships and ratios.

[0012] [Embodiment] Hereinafter, an overview of the system in the embodiment will be explained using the drawings. <System overview> First, an example overview of an information processing system 1 according to an embodiment will be described using FIG. 1. In the information processing system 1, for example, a user who measures brain waves wears an earphone set 10, in which a bioelectrode is provided in the ear canal, as a brain wave measurement device. In the example shown in FIG. 1, a neck-type earphone set 10 is used, but any earphone may be used as long as it is capable of sensing brain wave signals from the ear canal. For example, an earphone set that acquires a reference signal from the earlobe, an earphone that acquires a reference signal or a ground signal from another position (another position in the ear canal), or completely wireless earphones can be used. Furthermore, the electroencephalogram measurement device is not limited to the earphone type, but may be of the headphone type, for example, and may have a configuration in which a bioelectrode is provided in the earmuff portion. Further, the brain wave measuring device may be a device that measures brain waves from a part of the head other than the ear.

[0013] In the example shown in FIG. 1, the earphone set 10 acquires an electroencephalogram signal from the ear canal and transmits the electroencephalogram signal to the information processing device 30 or the information processing device 50 via the network N. The network N includes a wired or wireless network, and short-range wireless communication such as Bluetooth (registered trademark) may be used.

[0014] Furthermore, the earphone set 10 may perform predetermined processing on the brain wave signal and transmit it to the information processing device 30 that plays the role of a server or the information processing device 50 used by the user. The predetermined processing includes, for example, at least one of processing such as amplification processing, sampling, filtering, and differential calculation.

[0015] The information processing device 30 is, for example, a server, and sequentially acquires electroencephalogram signals measured by an electroencephalogram measurement device and executes each process. For example, the information processing device 30 sequentially acquires brain wave signals of a predetermined user stimulated by predetermined content, inputs the sequentially obtained brain wave signals into a brain wave model, and calculates the degree of arousal of the predetermined emotion in the predetermined content. Predict and predict the content of a given content. The content model is updated using the correspondence between feature values ​​and arousal degrees as relearning data, and content that arouses a predetermined emotion in a given user is extracted from the content list based on the arousal degree predicted by the content model. You can. This makes it possible to extract content that evokes a predetermined emotion in the user with high precision.

[0016] Here, the content may be anything that can stimulate the user, especially the user's five senses, and may be, for example, audio or music, photographs or videos, or things that emit scents. Note that although the content will be described as music below, the content is not limited to this.

[0017] Further, stimulating the user with content includes, for example, allowing the user to use the content. If the content is music, stimulating the user with the content includes, for example, having the user listen to the content. Note that the method of stimulating the user with content is not limited to this.

[0018] Further, the predetermined emotion may be any emotion, such as joy, anger, sadness, fun, love, sadness, loneliness, and nostalgia.

[0019] The information processing device 50 is, for example, a processing terminal such as a mobile terminal held by a user, and sequentially acquires brain wave signals from the earphone set 10. The information processing device 50 may, for example, use the earphone set 10 to play content to the user and sequentially acquire brain wave signals of the user who is stimulated by the content.

[0020] Furthermore, the information processing device 50 can obtain an evaluation from a user who has been stimulated by the content regarding whether or not a predetermined emotion has been aroused.

[0021] Details of the processing in the information processing device 30 and the information processing device 50 will be described later.

[0022] <Earphone set configuration> An overview of the earphone set 10 in the embodiment will be described using FIGS. 2 and 3. Note that the earphone set 10 is not limited to the examples shown in FIGS. 2 to 3, and any earphone can be applied to the technology of the present disclosure as long as it is capable of sensing brain waves from the ear canal and outputting them to an external device. I can do that.

[0023] FIG. 2 is a diagram showing an example of the earphone set 10 according to the embodiment. Earphone set 10 shown in FIG. 2 includes a pair of earphones 100R and 100L and a neck hanging portion 110. Each earphone 100R, 100L is connected to the neck hanging part 110 using a cable capable of signal communication, but may be connected using wireless communication. Hereinafter, RL will be omitted if there is no need to distinguish between left and right.

[0024] The neck hanging part 110 includes a central member that extends along the back of the neck, and rod-shaped members (arms) 112R and 112L that are curved along both sides of the neck. Electrodes 122, 124 for sensing brain wave signals are provided on the surface of the central member that contacts the back of the neck. Each of the electrodes 122 and 124 is a grounded electrode and a reference electrode. As a result, as will be described later, it is possible to increase the distance from the elastic electrode provided on the ear tip of the earphone, making it possible to acquire brain wave signals with high accuracy. Furthermore, the neck hanging unit 110 may include a processing unit that processes brain wave signals and a communication device that communicates with the outside, but these processing units and communication units may be provided in the earphone 100.

[0025] In addition, the rod-like members 112R and 112L on both sides of the neck hanging part 110 are heavier at their tips than at their bases (center member side), so that the electrodes 122 and 124 can be placed properly on the wearer's neck. It will become crimped. For example, weights are provided on the tip sides of the rod-shaped members 112R and 112L. Note that the positions of the electrodes 122 and 124 are not limited to these positions.

[0026] FIG. 3 is a diagram showing an example of a schematic cross section of the earphone 100R according to the embodiment. In the earphone 100R shown in FIG. 3, for example, an elastic member 108 (eg, urethane) may be provided between the speaker 102 and the nozzle 104. By providing this elastic member 108, vibrations of the speaker 102 are less likely to be transmitted to the elastic electrode of the ear tip 106, and it is possible to prevent sound interference between the elastic electrode of the ear tip 106 and the speaker 102.

[0027] Furthermore, although the ear tip 106 including the elastic electrode is located at the sound guide port, the elasticity of the elastic electrode itself makes it possible to prevent interference due to sound vibrations. Furthermore, by employing an elastic member for the housing, this elastic member makes it difficult for sound vibrations to be transmitted to the elastic electrodes of the ear tip 106, making it possible to prevent interference due to sound vibrations.

[0028] Earphone 100 includes an audio sound processor, and may use this audio sound processor to cut sound signals below a predetermined frequency (for example, 50 Hz) that corresponds to brain wave signals. In particular, audio sound processors cut sound signals below 30Hz, a frequency band that tends to show characteristics as brain wave signals, but they can also amplify sound signals at frequencies around 70Hz in order to avoid damaging the bass sound. good.

[0029] Thereby, interference between the sound signal and the brain wave signal can be prevented. In addition, the audio sound processor only needs to cut a predetermined frequency when sensing an electroencephalogram signal.

[0030] Furthermore, the ear tip 106 conducts an electroencephalogram signal sensed from the ear canal to a contact point of an electrode provided on the nozzle 104. The brain wave signal is transmitted from the ear tip 106 to a biological sensor (not shown) inside the earphone 100 via a contact point. The biosensor outputs the acquired brain wave signals to a processing device provided in the neck hanging part 110 or transmits them to an external device via a cable. Additionally, the ear tip 106 and the housing containing the biosensor and audio sound processor may be insulated. Note that, as described above, the electroencephalogram measurement device is not limited to the earphone set 10.

[0031] <Server configuration example> FIG. 4 is a block diagram showing an example of the information processing device 30 according to the embodiment. The information processing device 30 is, for example, a server, and may be composed of one or more devices. Further, the information processing device 30 processes the brain wave signal or brain wave information, and predicts, for example, the degree of arousal of a predetermined emotion due to the predetermined content.

[0032] The information processing device 30 is also referred to as a server 30. Note that the information processing device 30 does not necessarily have to be a server, and may be a general-purpose computer.

[0033] The server 30 includes, for example, one or more processors (CPU: Central Processing Unit) 310, one or more network communication interfaces 320, a memory 330, a user interface 350, and a system for interconnecting these components. Includes one or more communication buses 370.

[0034] Server 30 may optionally include a user interface 350, for example. User interface 350 may be, for example, a display device (not shown) and an input device such as a keyboard and / or mouse (or some other pointing device, not shown).

[0035] Memory 330 is, for example, a high speed random access memory such as DRAM, SRAM, DDR RAM or other random access solid state storage, and may also include one or more magnetic disk storage, optical disk storage, flash memory devices, or It may also be a nonvolatile memory such as another nonvolatile solid state storage device. Furthermore, the memory 330 may be a computer-readable non-transitory recording medium in which a program is recorded.

[0036] Also, another example of memory 330 may be one or more storage devices located remotely from processor 310. In some embodiments, memory 330 stores the following programs, modules and data structures, or a subset thereof.

[0037] One or more processors 310 read programs from memory 330 as needed and execute them. For example, by executing a program stored in the memory 330, one or more processors 310 can control the brain wave control unit 311, the brain wave acquisition unit 312, the first learning unit 313, the second learning unit 314, and the calculation unit 315. , an extraction unit 316, an update unit 317, a generation unit 318, and an output unit 319. The electroencephalogram control unit 311 controls and processes the sequentially acquired electroencephalogram signals, and controls the following processes.

[0038] The brain wave acquisition unit 312 sequentially acquires brain wave signals of a predetermined user stimulated by predetermined content from an electroencephalogram measurement device worn by the predetermined user. The predetermined content may be, for example, a predetermined content extracted by the extraction unit 316 described later, or may be selected content and non-selected content described later.

[0039] For example, the brain wave acquisition unit 312 sequentially acquires brain wave signals measured in the ear canal of a predetermined user from an electroencephalogram measurement device attached to the ear of the predetermined user. Note that the electroencephalogram measurement device is not limited to the earphone set 10. The electroencephalogram signal may be, for example, an electroencephalogram signal indicating feature data (power vector for each frequency per unit time) of the target user's in-ear EEG.

[0040] The first learning unit 313 inputs the feature amount of the content and generates a content model that predicts the degree of arousal of a predetermined emotion in a predetermined user. The content model is a learning model that inputs feature amounts of content and outputs a predicted degree of triggering. Hereinafter, the degree of attraction predicted by the content model will be referred to as the first degree of attraction.

[0041] When the content is music, the feature amount of the content may be a feature amount that represents 10 seconds of sound in 128 dimensions, for example.

[0042] Further, the first arousal degree may be, for example, the likelihood or probability that the content arouses a predetermined emotion in a predetermined user. The likelihood or probability may be a value of 0 to 1, for example.

[0043] The first learning unit 313, for example, calculates the feature amount of at least one selected content selected by a predetermined user as content that evokes a predetermined emotion, and the evaluation regarding the predetermined emotion by the predetermined user stimulated by the at least one selected content. The content is learned based on the correspondence relationship and the correspondence relationship between at least one non-selected content that has not been selected by a predetermined user and an evaluation regarding a predetermined emotion by a predetermined user who is stimulated by the at least one non-selected content. A model can be generated.

[0044] Specifically, for example, first, a predetermined user selects three contents as content that evokes a predetermined emotion. Further, for example, the information processing system 1 extracts three non-selected contents. At this time, the information processing system 1 may, for example, extract content having a feature amount different from the feature amount of the selected content as a non-selected content based on the feature amount of the content. Subsequently, the predetermined user is stimulated, for example, made to listen to the three selected contents and the three non-selected contents.

[0045] Then, an evaluation is obtained from the predetermined user regarding whether or not the content arouses a predetermined emotion. For example, evaluation may be performed by having a predetermined user answer a predetermined questionnaire. Note that the user may be asked to answer a questionnaire after being stimulated by each content, or may be made to answer a questionnaire after being stimulated by a plurality of contents. Here, when the content is music, the evaluation may be an evaluation based on heart rate variability analysis that analyzes the heartbeat of a predetermined user, or may be a subjective evaluation using a VAS (Visual Analog Scale).

[0046] After that, the first learning unit 313 generates a content model based on the correspondence between the feature amounts of the three selected contents and the three non-selected contents and their respective evaluations.

[0047] FIG. 5A is a diagram schematically showing a content model. The content model is generated, for example, by learning the correspondence between selected content or non-selected content (501) and evaluations regarding predetermined emotions (502).

[0048] In addition, the first learning unit 313 can update the content model by relearning, as relearning data, the correspondence between the feature amount of the predetermined content and the degree of arousal predicted by the electroencephalogram model described later. That is, the first learning unit 313 uses the correspondence relationship between the feature amount of the predetermined content and the degree of arousal predicted by the electroencephalogram model described later as training data for relearning the content model. Details of the relearning process will be described later.

[0049] The second learning unit 314 inputs an electroencephalogram signal of a predetermined user stimulated by the content and generates an electroencephalogram model that predicts the degree to which the content being stimulated will induce a predetermined emotion in the predetermined user. The electroencephalogram model is an electroencephalogram model that inputs an electroencephalogram signal and outputs a predicted degree of induction. Hereinafter, the induction degree output by the brain wave model will be referred to as the second induction degree.

[0050] The second arousal degree may be, for example, the likelihood or probability that the content arouses a predetermined emotion in a predetermined user. The likelihood or probability may be a value of 0 to 1, for example.

[0051] Here, the brain wave signals of the predetermined user stimulated by the content are, for example, the brain wave signals of the predetermined user who is stimulating the content, which are sequentially acquired from the earphone set 10.

[0052] The second learning unit 314, for example, determines the correspondence between an electroencephalogram signal of a predetermined user stimulated by at least one selected content and a label related to a predetermined emotion corresponding to the at least one selected content, and An electroencephalogram model can be generated based on the correspondence between an electroencephalogram signal of a predetermined user stimulated by the predetermined user and a label related to a predetermined emotion corresponding to at least one non-selected content.

[0053] Here, since the selected content is content selected by a predetermined user as content that evokes a predetermined emotion, it is, so to speak, content that corresponds to the correct label, and the second elicitation degree as the likelihood can be said to be "1". On the other hand, since the unselected content is content that has not been selected by the predetermined user, it is the content that corresponds to the incorrect label, so to speak, and the second triggering degree as the likelihood can be said to be "0". Therefore, the second learning unit 314 determines, for example, the correspondence between the electroencephalogram signal of a predetermined user stimulated by the selected content and the correct label, for example, the likelihood "1", and the correspondence between the electroencephalogram signal of the predetermined user stimulated by the non-selected content. An electroencephalogram model is generated by learning based on the correspondence between electroencephalogram signals and incorrect labels, such as likelihood "0".

[0054] Specifically, for example, first, a predetermined user selects three pieces of content as content that evokes a predetermined emotion. Further, for example, the information processing system 1 extracts three non-selected contents. Subsequently, the predetermined user is stimulated, for example, made to listen to the three selected contents and the three non-selected contents.

[0055] Then, the brain wave acquisition unit 312 sequentially acquires brain wave signals of a predetermined user who is stimulating each content. Thereafter, the second learning unit 314 generates an electroencephalogram model based on the correspondence between the electroencephalogram signals during stimulation of the three selected contents and the three non-selected contents and their respective labels.

[0056] FIG. 5B is a diagram schematically showing an electroencephalogram model. The brain wave model is generated, for example, by learning the correspondence between the brain waves of the selected content or the brain waves of the non-selected content (503) and a label (504) related to a predetermined emotion.

[0057] The calculation unit 315 calculates an attraction score for each of the plurality of candidate contents based on each of the degrees of attraction predicted by inputting the feature amounts of each of the plurality of candidate contents in the content list into the content model.

[0058] The arousal score is a score that is evaluated and calculated through the information processing system 1 and corresponds to the degree to which the content arouses a predetermined emotion in a predetermined user. Note that the arousal score may be the same value as the first arousal degree, but preferably, as will be described later, content that arouses a predetermined emotion in a predetermined user is included in teacher data when learning a content model. and the candidate content in the content list.

[0059] The calculation unit 315 calculates an arousal score based on the degree of similarity between each of the candidate contents in the content list and the content that arouses a predetermined emotion in a predetermined user, which is included in the teacher data used when learning the content model. You can. Thereby, the information processing device 30 can calculate the elicitation score by considering both the content model and the electroencephalogram model, and can improve the accuracy of the content model through repeated processing through relearning processing, which will be described later. . Note that details of the repetitive processing will be described later.

[0060] Here, the content that arouses a predetermined emotion in a predetermined user and is included in the teacher data during content model learning may be, for example, selected content. Specifically, the calculation unit 315 calculates an attraction score based on the degree of similarity between each of the candidate contents in the content list and the selected contents initially selected by the user, for example, three selected contents. Good too.

[0061] Moreover, the content that arouses a predetermined emotion in a predetermined user, which is included in the teacher data during content model learning, may be, for example, high-ranking content in an updated content list that will be described later. Specifically, the calculation unit 315 may calculate the attraction score, for example, based on the degree of similarity between each of the candidate contents in the content list and a predetermined number of high-ranking contents in the content list.

[0062] Here, in the content list, for example, the content is associated with information indicating the order of the degree to which the content arouses a predetermined emotion in a predetermined user. The degree to which a predetermined emotion is evoked in a predetermined user may be a first degree of arousal, or may be an arousal score calculated in advance, for example, by the calculation unit 315 based on a content list before being updated.

[0063] The calculation unit 315 calculates, for example, a first arousal degree as a likelihood having a value of 0 to 1, and a similarity degree as a correlation coefficient between content that evokes a predetermined emotion in a predetermined user and candidate content in the content list. The attraction score can be calculated by performing predetermined calculation processing, for example, addition processing.

[0064] Further, the calculation unit 315 can repeat calculation of the attraction score through an iterative process described later.

[0065] The extraction unit 316 extracts content that arouses a predetermined emotion in a predetermined user from the content list based on the first degree of arousal.

[0066] Further, the extraction unit 316 extracts a predetermined content that arouses a predetermined emotion in a predetermined user from among the plurality of candidate contents based on the calculated arousal score.

[0067] At this time, the extraction unit 316 may extract at least one of the predetermined content having a first attraction degree or attraction score equal to or greater than a predetermined value, and the predetermined content having a first attraction degree or attraction score less than or equal to a predetermined value. can. Thereby, the information processing device 30 can extract at least one of the predetermined content that causes a predetermined emotion and the predetermined content that does not cause a predetermined emotion, and can, for example, efficiently perform a relearning process. , it is possible to effectively stimulate a predetermined user using content that arouses a predetermined emotion in the predetermined user.

[0068] Furthermore, the extraction unit 316 can re-extract the predetermined content from among the plurality of candidate contents based on the attraction score that is recalculated through an iterative process that will be described later. Furthermore, the extraction unit 316 can repeatedly extract the predetermined content through an iterative process that will be described later. Here, the predetermined content to be re-extracted may be any content, and may be content that has stimulated the user in the past, or content that has not stimulated the user.

[0069] The updating unit 317 updates the content list based on the attraction score.

[0070] The updating unit 317 can update the content list by rearranging the order of the contents included in the content list, for example, in order of attraction score. Thereby, the information processing system 1 can manage the contents in association with the order in which they evoke a predetermined emotion in a predetermined user.

[0071] The updating unit 317 can repeatedly update the content list through an iterative process that will be described later.

[0072] The generation unit 318 identifies the feature amount of the content that evokes a predetermined emotion in a predetermined user based on the content model, and generates predetermined content that stimulates the predetermined user based on the feature amount.

[0073] Furthermore, the generation unit 318 may generate the predetermined content based on the updated content model.

[0074] The output unit 319 outputs a content list.

[0075] The output unit 319 can output, for example, a content list in which the order of attraction score and content are associated with each other.

[0076] The output unit 319 can output the content list to the information processing device 50, for example.

[0077] Further, the output unit 319 may output a content list including the content generated by the generation unit 318.

[0078] Through the above processing, the information processing device 30 can update the content model using the output of the electroencephalogram model as relearning data for the content model, and extract the predetermined content that evokes the predetermined emotion in the predetermined user. Thereby, the information processing device 30 can extract content that evokes a predetermined emotion in a predetermined user with high precision.

[0079] <Example of configuration of processing terminal> FIG. 6 is a block diagram showing an example of the information processing device 50 according to the embodiment. The information processing device 50 includes, for example, a medical device, a mobile terminal (such as a smartphone), a computer, a tablet terminal, etc., as described above. The information processing device 50 is also referred to as a processing terminal 50.

[0080] Processing terminal 50 may include, for example, one or more processors (e.g., CPUs) 510, one or more network communication interfaces 520, memory 530, user interface 550, and one for interconnecting these components. or includes multiple communication buses 570.

[0081] User interface 550 includes, for example, a display 551 and an input device 552 (such as a keyboard and / or mouse or some other pointing device). Further, the user interface 550 may be a touch panel.

[0082] Memory 530 is, for example, a high speed random access memory such as DRAM, SRAM, DDR RAM or other random access solid state storage, and may also include one or more magnetic disk storage, optical disk storage, flash memory devices, or It may also be a nonvolatile memory such as another nonvolatile solid state storage device. Furthermore, the memory 530 may be a computer-readable non-transitory recording medium that records a program.

[0083] Additionally, other examples of memory 530 may include one or more storage devices located remotely from processor 510. In some embodiments, memory 530 stores the following programs, modules and data structures, or a subset thereof.

[0084] One or more processors 510 read and execute programs from memory 530 as needed. For example, one or more processors 510 may configure an application control unit (hereinafter also referred to as “application control unit”) 511 by executing a program stored in memory 530. The application control unit 511 is an application that processes brain wave signals, and includes, for example, a selection unit 512, an evaluation acquisition unit 513, and an output unit 514.

[0085] The selection unit 512 selects content in the content list as selected content based on a predetermined user's operation.

[0086] The evaluation acquisition unit 513 acquires an evaluation from a predetermined user regarding whether a predetermined emotion is aroused by the content. For example, the processing terminal 50 may display a screen for accepting responses to a predetermined questionnaire, and the evaluation acquisition unit 513 may obtain the predetermined user's response based on the predetermined user's operation on the screen. When the content is music, the evaluation acquisition unit 513 may acquire the results of an evaluation based on heart rate variability analysis that analyzes the heartbeat of a predetermined user, or a subjective evaluation using a VAS (Visual Analog Scale).

[0087] The output unit 514 outputs, to the information processing device 30, the result of the evaluation regarding whether or not the content arouses a predetermined emotion, which has been acquired by the evaluation acquisition unit 513.

[0088] Through the above processing, the information processing device 50 can output to the information processing device 30 an evaluation as to whether a predetermined emotion is aroused by the selected content and non-selected content, and the information processing device 30 can output the evaluation based on the evaluation. content can be generated.

[0089] <Operation> Next, the operation according to the embodiment will be explained. FIG. 7 is a flowchart illustrating an example of processing in the information processing system 1. The flowchart shown in FIG. 7 shows an example of content model generation processing in the information processing system 1 according to the embodiment.

[0090] The selection unit 512 selects content in the content list as selected content based on a predetermined user's operation (S702). The evaluation acquisition unit 513 acquires evaluations regarding whether a predetermined emotion is aroused by the selected content and non-selected content (S704). The output unit 514 outputs the evaluation result to the information processing device 30 (S706).

[0091] The first learning unit 313 generates a content model based on the correspondence between the feature values ​​of the selected content and the evaluation regarding the predetermined emotion, and the correspondence between the feature values ​​of the non-selected content and the evaluation regarding the predetermined emotion (S708 ).

[0092] FIG. 8 is a flowchart illustrating an example of processing in the information processing system 1. The flowchart shown in FIG. 8 shows an example of the electroencephalogram model generation process in the information processing system 1 according to the embodiment.

[0093] For example, the earphone set 10 stimulates a predetermined user using content (S802). The brain wave acquisition unit 312 sequentially acquires brain wave signals of a predetermined user stimulated by each of the selected content and non-selected content (S804). The second learning unit 314 determines the correspondence between a predetermined user's brain wave signal stimulated by selected content and a label related to a predetermined emotion corresponding to the selected content, and a predetermined user's brain wave signal stimulated by non-selected content. An electroencephalogram model is generated based on the correspondence with the label related to the predetermined emotion corresponding to the non-selected content (S806).

[0094] FIG. 9 is a flowchart showing an example of processing in the information processing system 1. The flowchart shown in FIG. 9 shows an example of content model relearning processing in the information processing system 1 according to the embodiment.

[0095] The extraction unit 316 extracts predetermined content from the content list based on the attraction score (S902). The information processing device 30 stimulates a predetermined user with predetermined content, for example, reproduces the predetermined content (S904), and sequentially acquires brain wave signals of the predetermined user (S906). The information processing device 30 inputs the acquired electroencephalogram signal into the electroencephalogram model (S908) and predicts the degree of induction (S910).

[0096] Then, the information processing device 30 (in particular, the first learning unit 313) relearns the correspondence between the extracted predetermined content and the degree of arousal predicted by the electroencephalogram model as relearning data (S912), and uses the content model as relearning data. Update (S914).

[0097] FIG. 10 is a flowchart illustrating an example of processing in the information processing system 1. The flowchart shown in FIG. 10 shows an example of content list update processing in the information processing system 1 according to the embodiment.

[0098] First, the information processing device 30 inputs the feature amount of each content included in the content list into the content model (S1002), and predicts the degree of triggering of each content (S1004). Furthermore, the information processing device 30 calculates the degree of similarity between each content included in the content list and the selected content (S1006). The information processing device 30 (particularly the calculation unit 315) calculates an attraction score based on the predicted degree of attraction and the degree of similarity (S1008). The information processing device 30 updates the content list (S1010).

[0099] Then, the information processing device 30 performs an iterative process of repeating the content model update process and the content list update process. Thereby, the information processing system 1 can improve the accuracy of the content model, and can also highly accurately extract content that evokes a predetermined emotion in a predetermined user based on the updated content list.

[0100] Note that in the iterative process, the content model update process and the content list update process may be performed alternately once at a time. Thereby, the information processing system 1 can more accurately extract content that evokes a predetermined emotion through simple iterative processing.

[0101] Furthermore, in the iterative process, for example, the content model update process may be performed after the content list update process is performed multiple times, or the content list update process may be performed after the content model update process is performed multiple times. You may do so. Thereby, the information processing system 1 can flexibly execute repetitive processing depending on the amount of calculation and required accuracy.

[0102] Furthermore, the information processing device 30 can specify the feature amount of the content that evokes a predetermined emotion based on the updated content model, and can generate the predetermined content that stimulates the predetermined user. Thereby, the information processing device 30 can newly generate predetermined content that arouses a predetermined emotion in a predetermined user.

[0103] Furthermore, the information processing device 30 can include the generated predetermined content in the content list. Thereby, the information processing device 30 can allow a predetermined user to use, for example, listen to, the generated predetermined content.

[0104] Note that this embodiment is provided to facilitate understanding of the present invention, and is not intended to be interpreted as limiting the present invention. The present invention may be modified / improved without departing from its spirit, and the present invention also includes equivalents thereof.

[0105] Furthermore, in the present invention, a "section" does not simply mean a physical means, but also includes a case where the function of the "section" is realized by software. Furthermore, even if the functions of one "part" or device are realized by two or more physical means, devices, or software, the functions of two or more "parts" or devices are realized by one physical means, device, or software. , or may be realized by software. [Explanation of symbols]

[0106] 1 Information processing system 10 earphone set 30, 50 Information processing equipment 100 earphones 102 Speaker 104 nozzle 106 ear tips 310 processor 311 Brain wave control section 312 EEG acquisition section 313 1st Learning Department 314 2nd Learning Department 315 Calculation part 316 Extraction part 317 Update Department 318 Generation part 319 Output section 330 memory 510 processor 511 Application control section 512 Selection section 513 Evaluation Acquisition Department 514 Output section 530 memory

Claims

1. The processor included in the information processing device The process involves sequentially acquiring the brainwave signals of a designated user stimulated by a designated content from an electroencephalogram (EEG) measurement device attached to the designated user, The brainwave signals acquired in the order described above are input to an electroencephalogram (EEG) model that receives the brainwave signals of a predetermined user stimulated by the content and predicts the degree to which the content stimulates the predetermined user to evoke a predetermined emotion, thereby obtaining the degree to which the predetermined emotion is evoked by the predetermined content as predicted by the EEG model. The process involves inputting the content features into a content model that has learned the correspondence between the features of the predetermined content and the degree of arousal of the predetermined emotion for the predetermined user, including the predicted degree of arousal, in order to obtain the degree of arousal predicted by the content model. From the content list, extract predetermined content that evokes a predetermined emotion in a predetermined user, based on the degree of relevance predicted by the content model. A program that executes the command.

2. The aforementioned processor, The feature quantities of each of the multiple candidate contents in the content list are input into the content model, and based on the predicted triggering degree of each of the multiple candidate contents, the triggering score for each of the multiple candidate contents is calculated. The extraction includes extracting predetermined content from among the plurality of candidate content that evokes the predetermined emotion of the predetermined user, based on the calculated trigger score. Acquiring in the order described above includes sequentially acquiring the brainwave signals of the predetermined user being stimulated by the predetermined content extracted, The program according to claim 1.

3. The program according to claim 2, wherein the extraction includes extracting at least one of the predetermined content having an elicitation score greater than or equal to a predetermined value and the predetermined content having an elicitation score less than or equal to a predetermined value.

4. The program according to claim 2, wherein the calculation further includes calculating the arousal score based on the similarity between the content that evokes a predetermined emotion of a predetermined user, which is included in the training data used when learning the content model, and candidate content in the content list.

5. The program according to any one of claims 1 to 4, wherein the content model is generated based on the correspondence between the feature quantities of at least one selected content chosen by the predetermined user as content that evokes the predetermined emotion and the evaluation of the predetermined emotion by the predetermined user stimulated by the at least one selected content, and the correspondence between at least one non-selected content not chosen by the predetermined user and the evaluation of the predetermined emotion by the predetermined user stimulated by the at least one non-selected content.

6. The program according to claim 5, wherein the electroencephalogram model is generated based on the correspondence between the electroencephalogram signals of a predetermined user stimulated by the at least one selected content and a label relating to a predetermined emotion corresponding to the at least one selected content, and the correspondence between the electroencephalogram signals of a predetermined user stimulated by the at least one non-selected content and a label relating to a predetermined emotion corresponding to the at least one non-selected content.

7. An information processing device including a processor, The aforementioned processor, The process involves sequentially acquiring the brainwave signals of a designated user stimulated by a designated content from an electroencephalogram (EEG) measurement device attached to the designated user, The brainwave signals acquired in the order described above are input to an electroencephalogram (EEG) model that receives the brainwave signals of a predetermined user stimulated by the content and predicts the degree to which the content stimulates the predetermined user to evoke a predetermined emotion, thereby obtaining the degree to which the predetermined emotion is evoked by the predetermined content as predicted by the EEG model. The process involves inputting the content features into a content model that has learned the correspondence between the features of the predetermined content and the degree of arousal of the predetermined emotion for the predetermined user, including the predicted degree of arousal, in order to obtain the degree of arousal predicted by the content model. From the content list, extract predetermined content that evokes a predetermined emotion in a predetermined user, based on the degree of relevance predicted by the content model. An information processing device that performs the following actions.

8. The processor included in the information processing device The process involves sequentially acquiring the brainwave signals of a designated user stimulated by a designated content from an electroencephalogram (EEG) measurement device attached to the designated user, The brainwave signals acquired in the order described above are input to an electroencephalogram (EEG) model that receives the brainwave signals of a predetermined user stimulated by the content and predicts the degree to which the content stimulates the predetermined user to evoke a predetermined emotion, thereby obtaining the degree to which the predetermined emotion is evoked by the predetermined content as predicted by the EEG model. The process involves inputting the content features into a content model that has learned the correspondence between the features of the predetermined content and the degree of arousal of the predetermined emotion for the predetermined user, including the predicted degree of arousal, in order to obtain the degree of arousal predicted by the content model. From the content list, extract predetermined content that evokes a predetermined emotion in a predetermined user, based on the degree of relevance predicted by the content model. An information processing method that performs the following.

Images