JP2010237729A - Subjective evaluation value detection device, subjective evaluation value detection method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a subjective evaluation value detection device for evaluating the degree of subjective view of a person from information on switch operation based on his or her sense. <P>SOLUTION: The switch 101 in intuitively operated by a user subjectively for a predetermined sense. A feature amount extraction unit 102 extracts input operation information that is a set of predetermined feature amounts from the input operation to the switch 101. An input operation information storage unit 103 accumulates the input operation information. An input operation profile information generation unit 104 determines an evaluation axis where the average information amount (entropy) of the combination of predetermined feature amounts of the user's input operation based on the input operation information. An evaluation value calculation unit 106 calculates an evaluation value by projecting the input operation information to the evaluation axis determined by the input operation profile information generation unit 104. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a subjective rating value detection apparatus, a subjective rating value detection method, and a program for rating the degree of subjectivity based on information that a person operates a switch according to his / her sense. More specifically, the present invention relates to subjective rating value detection that detects a rating value of the degree of flashing that occurs in human activity.

  A flash is a phenomenon in which wonderful thoughts come to mind instantly. When the flash occurs, the nerve cells are activated all at once in the brain for a short time of about 0.1 seconds, and try to stop the inspiration. It is also known that at that moment, the A10 nerve is activated and dopamine is released, and pleasure is obtained. However, since the activity of nerve cells that stops flashing is a momentary fixing action, learning by nerve cells in the brain is not sufficient, and the inspirational contents cannot be memorized and are often forgotten immediately.

  As a method of collecting subjective information including the degree of flashing, etc., a method of selecting a graded method called a four-method method or a five-method method, or a position representing an appropriate degree from the range of a meter-type rating is determined. A grading method may be used.

  Regarding the flash rating method using a web page, the surveyor creates a web page including rating method options or meters, the respondent answers via the network, the input answer is stored in the server, and the server There are many things that can automatically perform processing such as summing up the accumulated answers.

  In particular, there are those that collect response data from participants in response to questionnaires in real time in television programs, exhibitions or lectures. In relation to the collection of answer data, a system is disclosed that collects answers to questions in real time using a wireless input terminal having a plurality of buttons for inputting results of selecting answers to the questions (patent) Reference 1).

  The questionnaire data collection system of Patent Document 1 includes a switch unit having a plurality of switches and a transmission unit that wirelessly transmits operation information of each switch of the switch unit, and each transmission unit is different among a plurality of communication channels. A plurality of operation terminal devices that selectively transmit using a communication channel, a reception device that receives signals from each transmission unit of the plurality of operation terminal devices, and operation information of each switch received by the reception device And a collection processing means for performing a predetermined process for display.

With respect to speech recognition, Patent Document 2 describes a hierarchical eigenspace extraction method that is effective even when the number of adaptive data is small and that can avoid computational complexity. The technology of Patent Document 2 is
Speaker data for learning of a large number of speakers is input to a speaker model creating means to create a speaker model. The speaker model is input to the hierarchical class eigenspace extracting means, and eigenspaces from which knowledge based on the speaker characteristics of the nodes or leaves in the immediately lower hierarchy is extracted are extracted for each node.

  Patent Document 3 describes that a user's TV viewing channel transition history is stored, and a user profile indicating an operation pattern related to the user's channel (program) transition is created from the stored data.

  In addition, Patent Document 4 describes a technique for determining whether or not a person shown in two face images is the same person even when there is a change in posture, illumination, or the like for each registered person. . The technique of Patent Document 4 generates a plurality of variation images having different postures, face positions, and sizes from a normalized image, and extracts frequency features from the plurality of variation images. Further, the frequency feature is projected onto a discriminant space having high discriminability obtained by linear discriminant analysis. The standard person comparison means performs standard person comparison and extracts features with high discriminability. On the other hand, with respect to the collation image, the discrimination feature is extracted using the feature extraction unit and the discrimination space projection unit. The score calculation means outputs a collation score using the discrimination axis obtained from the registered image and the discrimination feature obtained from the collation image. The collation determining unit determines whether or not the same person is present by comparing the collation score with a threshold value.

Japanese Patent Laid-Open No. 11-039351 JP 2003-216178 A JP-A-2005-130087 JP 2006-338092 A

  Regarding the collection of the degree of flashing, the first problem that arises in the rating method using related technical questionnaires and web pages is the stage and range in which the rating value of the answer is set in advance. However, it is required to convert the flashing level into a rating value in consideration of future events, and an appropriate rating value cannot be answered. The second problem is that respondents need to check the questions and grades displayed on the paper or on the web page, select or describe them, and answer them before forgetting the flash that occurred instantaneously. In daily life, every time a flash occurs, immediately confirming the display and giving an answer is a point that causes a heavy burden on the respondent's life. The third problem is that an accurate rating value cannot be collected due to the ambiguity of memory in the method of collecting responses by looking back on past events at regular intervals such as several days.

On the other hand, in Patent Document 1, although rating values can be collected from a plurality of respondents in real time, since the rating value of the answer, which is the first problem described above, is a preset stage and range, It is required to respond with an appropriate rating value instantly in consideration of
Appropriate rating values cannot be answered.

  For example, when the rating is set in five stages in advance and the respondent responds that the event is a flash of grade “5” in an event, the respondent is different from the event in the subsequent event. It is not possible to answer with an appropriate value for a higher degree event.

  The present invention has been made in view of the circumstances as described above, and an object of the present invention is to provide a subjective rating value detection device that evaluates the degree of subjectivity from information that a person operates a switch according to the sense.

In order to achieve the above object, a subjective rating value detection apparatus according to the first aspect of the present invention provides:
A switch that allows a user to intuitively input a predetermined sense according to its subjectivity,
Feature extraction means for extracting a set of the predetermined feature values from the input operation of the switch;
Storage means for storing the set of feature quantities extracted by the feature extraction means;
Profile generation means for obtaining an evaluation axis that maximizes the average information amount (entropy) of the predetermined feature quantity set of the input operation of the user based on the feature quantity set accumulated in the storage means;
Rating means for calculating a rating value obtained by projecting the set of predetermined feature values extracted from one of the input operations onto the evaluation axis obtained by the profile generating means;
It is characterized by providing.

The subjective rating value detection method according to the second aspect of the present invention is:
A step of accepting an input for the user to operate the switch intuitively according to the subjectivity for a predetermined sense;
A feature extraction step of extracting the predetermined set of feature values from the input operation of the switch;
A storage step of storing the set of feature quantities extracted in the feature extraction step;
A profile generation step for obtaining an evaluation axis that maximizes an average information amount (entropy) of a predetermined feature amount set of the input operation of the user based on the feature amount set accumulated in the storage step;
A rating step for calculating a rating value obtained by projecting the set of predetermined feature values extracted from one of the input operations onto the evaluation axis obtained in the profile generation step;
It is characterized by providing.

A program according to a third aspect of the present invention is stored in a computer.
A step of accepting an input for the user to operate the switch intuitively according to the subjectivity for a predetermined sense;
A feature extraction step of extracting the predetermined set of feature values from the input operation of the switch;
A storage step of storing the set of feature quantities extracted in the feature extraction step;
A profile generation step for obtaining an evaluation axis that maximizes an average information amount (entropy) of a predetermined feature amount set of the input operation of the user based on the feature amount set accumulated in the storage step;
A rating step for calculating a rating value obtained by projecting the set of predetermined feature values extracted from one of the input operations onto the evaluation axis obtained in the profile generation step;
Is executed.

  According to the present invention, the rating value of the respondent's sense can be calculated from the information that the respondent intuitively operated the switch according to the subjectivity. As a result, it is possible to reduce the psychological burden of the respondent himself who converts the feeling into the rating value. The reason is that the rating value is automatically calculated from the analysis of the change in the input operation information without using a preset stage or range for inputting the answer.

It is a block diagram which shows the structural example of the subjective rating value detection apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the example of the operation information of the switch in embodiment of this invention. It is a figure which shows the example which extracts the feature-value from operation information. It is a figure explaining the example of the method of calculating a rating value from the feature-value of operation information. 4 is a flowchart illustrating an example of an operation for generating input operation profile information in the first embodiment. 3 is a flowchart showing an example of an operation of rating value calculation in the first embodiment. It is a block diagram which shows the structural example of the subjective rating value detection apparatus which concerns on Embodiment 2 of this invention. 10 is a flowchart illustrating an example of an operation of the answer input terminal device according to the second embodiment. 12 is a flowchart illustrating an example of an operation of generating input operation profile information in the second embodiment. It is a block diagram which shows the physical structural example of the subjective rating value detection apparatus which concerns on embodiment of this invention.

  EMBODIMENT OF THE INVENTION Below, the form for inventing is demonstrated in detail with reference to drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals.

(Embodiment 1)
FIG. 1 is a block diagram showing a configuration example of a subjective rating value detection apparatus according to Embodiment 1 of the present invention. The subjective rating value detection apparatus 10 according to the first exemplary embodiment includes a switch 101, a feature amount extraction unit 102, an input operation information storage unit 103, an input operation profile information generation unit 104, an input operation profile information storage unit 105, and a rating value calculation unit 106. And a rating value storage unit 107.

  These means generally operate as follows. The switch 101 is operated intuitively according to the subjectivity of a predetermined feeling of the respondent. The subjective rating value detection apparatus 10 can target various sensations. For example, it is intended for feelings that can operate the switch 101 based on, for example, inspiration, fun, fun, sadness, anger, pain, beauty, discomfort, and the like.

  In the first embodiment, the respondent operates to input the degree of flashing. Here, flashing is assumed to be the feeling that the respondent himself thought of having made a new discovery or invention. The respondent operates the switch 101 in accordance with the intuition that the flash thought to be, such as the size, depth, novelty, sharpness of the flash, the spread of the expected influence, and so on. The switch 101 is an operation unit that allows the respondent to intentionally change the operation intensity, duration, frequency, and the like. For example, a means capable of operating on / off binary information such as a push button or a means capable of operating a plurality of real value information such as an acceleration sensor may be used. As the switch 101, various types of switches 101 such as a push button, a momentary lever switch, a jog dial, a joystick, and an automatic return rotation type can be used. For example, the respondent operates the switch 101 according to the feeling of flashing, such as large, slow, fast and small increments, fast, large, quick repetition, and slow large repetition.

  The feature amount extraction unit 102 receives a signal generated when the switch 101 is operated, and extracts the feature amount of the switch operation. The feature amount of the switch operation includes, for example, the number of times the switch 101 is operated at an interval equal to or less than a predetermined time, the frequency of operating the switch 101 per predetermined time, the time that the switch 101 is operated and held, For example, the operation interval, the magnitude of the operation of the switch 101, the speed at which the switch 101 is operated, the magnitude of the force to operate the switch 101, or the magnitude of the moment to operate the switch 101. The feature amount extraction unit 102 extracts the amount of one or more sets of feature amounts as described above. Hereinafter, a set of feature amounts may be simply referred to as a feature amount. A set of feature values of switch operation is also referred to as input operation information for easy understanding.

  The feature amount extraction unit 102 sends the feature amount extracted from the operation of the switch 101 to the input operation information storage unit 103. The feature amount extraction unit 102 may add information about the time of input operation to the feature amount. The input operation information storage unit 103 sequentially stores (accumulates) the received operation information. Note that the waveform of the signal operated by the switch 101 may be stored as A / D converted data at an appropriate sampling period, for example.

  The input operation profile information generation unit 104 generates the input operation profile information of the respondent based on the input operation information (a set of feature amounts) accumulated in the input operation information storage unit 103. The input operation profile information is an evaluation axis that maximizes the average information amount (entropy) of the set of feature amounts of the input operation. In the case where the set of feature amounts is one-dimensional consisting of one feature amount, the evaluation axis is, for example, conversion that averages the density of the distribution of feature amounts or conversion that expands the distribution of feature amounts. When the feature set consists of two or more feature values, for example, the sparseness of the distribution of points obtained by dimension reduction and projection of the feature set is averaged, or the feature set is An evaluation axis that expands the variance of the projected points is obtained.

  For the dimension reduction, for example, a principal component analysis method can be used. The evaluation axis is not limited to one dimension, and may be two dimensions (plane) or three dimensions (solid). For example, in the principal component analysis method, the first principal component is typically used as the evaluation axis, but the second principal component and the third principal component may be added to the evaluation axis.

  The input operation profile information generation unit 104 sends the generated input operation profile information to the input operation profile information storage unit 105. The input operation profile information storage unit 105 stores the received input operation profile information. Note that a waveform of a signal operated by the switch 101 may be stored, and a set of feature amounts may be extracted from the waveform of the signal and used when generating input operation profile information.

  The rating value calculation unit 106 calculates the rating value of each input operation information stored in the input operation information storage unit 103 using the input operation profile information stored in the input operation profile information storage unit 105. The rating value is a value obtained by projecting the input operation information onto the evaluation axis. When the input operation information is one-dimensional, the rating value is, for example, a conversion that averages the density of the distribution of feature amounts. When the input operation information is two-dimensional or more, the rating value is a coordinate obtained by projecting a set of feature amounts for each input operation information onto the evaluation axis of the input operation profile information. In particular, when the dimension-reduced evaluation axis is one-dimensional, the rating value is one value.

  The rating value storage unit 107 stores the calculated rating value. The rating value may be stored in association with the time (input operation time) of the input operation information.

  The timing for generating the input operation profile information from the accumulated input operation information may be performed every time the input operation information is newly stored, or may be performed when the rating value is used externally. Similarly, the timing for calculating the rating value from the input operation information may be performed every time input operation information is newly stored, or may be simultaneously with the timing for generating the input operation profile information, or the profile information may be evaluated. It may be performed when the value is used externally.

  Based on the rating value calculated by the rating value calculation unit 106, the sense of the respondent when performing a focused input operation can be evaluated. For example, by displaying the distribution of evaluation values of input operation information and the evaluation values of interest in the graph, the position of the respondent's sense when performing the input operation of interest can be grasped.

  By associating the input operation information and the rating value with the time when the input operation was performed, other information such as the activity record of the respondent, the voice of the respondent, the video of the respondent, or the biometric information of the respondent is keyed. By doing so, it is possible to analyze the feelings of the respondents.

  The operation of the embodiment of the present invention will be described below using a specific example. As a specific example, a case will be described in which a flash evaluation value is automatically calculated using a push button terminal device connected to the switch 101 via a wireless network.

  When respondents feel a flash in their daily lives, they operate the push buttons intuitively according to the flash. A signal generated by operating the push button (switch 101) is sent to the feature amount extraction unit 102. In this example, as shown in FIG. 2, it is assumed that binary information on ON / OFF of the push button is transmitted.

  The feature quantity extraction unit 102 extracts a set of feature quantities from the signal input from the switch 101. FIG. 3 shows an example of the feature amount (input operation information) extracted from the signal when the switch 101 is operated. In FIG. 3, the number of times of pressing and the average pressing time are used as a set of feature amounts. When the operation information at a certain time point t1 is set as a starting point, input operation information that appears at an interval equal to or less than a preset input operation time interval β is regarded as one continuous operation. If the number of times of pressing is N and the time of pressing N times is P1, P2,..., Pn, the average pressing time can be calculated by (P1 + P2 +... + Pn) / N.

  The feature amount may be freely set depending on the device used for the switch 101. Also, in order to cope with various operations of respondents, it is desirable to prepare a large number of feature quantities that have independent meanings.

  The feature quantity extraction unit 102 sends the extracted feature quantity set to the input operation information storage unit 103. The input operation information storage unit 103 stores the input operation information sent from the feature amount extraction unit 102 together with time information associated with a time measuring means (not shown).

  The input operation profile information generation unit 104 processes the input operation information accumulated in the input operation information storage unit 103 at the time when a preset number of new input operation information is obtained, and generates input operation profile information. In the generation procedure of the input operation profile information, first, for each recorded input operation information, a value related to a single or several types of feature values Xi is extracted. As the feature amount, one that takes into consideration the intensity, duration, frequency, etc. of the operation is used. In this example, it is assumed that the number of times of pressing X1 and the average pressing time X2 in the continuous operation of the push button are calculated as feature amounts.

  The extracted set of feature quantities is normalized so that the average is 0 and the variance is 1 for the entire operation information. Thereafter, dimension reduction is performed on each normalized operation information. In this example, principal component analysis (PCA) is performed.

  The first principal component of the principal component vector that can be calculated as a result of the principal component analysis is an evaluation axis that maximizes the average information amount (entropy) of each input operation information as shown in FIG. In FIG. 4, the first principal component is indicated by a thick solid line. The direction of the arrow of the first principal component represents the positive direction of the rating value. The point where the thin solid line intersects with the thick solid line of the first main component at right angles indicates the reference point of the rating value. The first principal component is stored in the input operation profile information storage unit 105 as the input operation profile information of the respondent. The input operation profile information includes the relational expression (straight line) of the first principal component, the coordinates of the reference point (rating value = 0 point) on the straight line, and the unit length on the straight line, with the feature quantity set as the coordinate system That's it.

  The rating value calculation unit 106 calculates a rating value using the input operation profile information stored in the input operation profile information storage unit 105. As shown in FIG. 4, the rating value is calculated by projecting a vector having elements extracted from each input operation information as elements onto an evaluation axis stored as input operation profile information.

  Therefore, the average value of the user's average input operation information is zero. In the positive direction of the first principal component (in the direction of the arrow of the first principal component shown by the thick solid line in FIG. 4), the rating value becomes positive as the input operation has a greater number of pressings than the average and a shorter average pressing time. Input operation information located in the negative direction of the first principal component from the reference point is calculated as a negative value. The calculated rating value is stored in the rating value storage unit 107 together with the input operation time. Taking FIG. 3 as an example, the input operation time to be stored may be the start time t1 of the input operation information determined as a series of operations, or may be the middle of the input operation time (t1 + Tc_1 / 2).

  FIG. 5 is a flowchart illustrating an example of an operation for generating input operation profile information according to the first embodiment. Although not described in the flowchart, the input operation profile information generation sets the counter Ta to 0 when activated.

  The subjective rating value detection apparatus 10 waits for an input operation of the switch 101 to be performed (step S11, step S12; NO). When the respondent recognizes the flashing and intuitively operates the switch 101 with respect to the flashing degree (step S12; YES), the feature amount extraction unit 102 receives the input operation signal, and the predetermined feature amount is obtained from the input operation signal. Are extracted (input operation information) (step S13). At the same time, the time of the input operation is counted and associated with the input operation information (step S14).

  The feature quantity extraction unit 102 accumulates the extracted feature quantity sets (input operation information) in the input operation information storage unit 103 (step S15). The feature quantity extraction unit 102 adds 1 to the counter Ta (step S16). If Ta is less than the predetermined number of times α (step S17; NO), the process returns to step S11 and is repeated from the switch operation input. Thus, each time the switch 101 is operated, input operation information is extracted and stored in the input operation information storage unit 103 together with the operation time.

  When the new storage number Ta reaches the predetermined number α (step S17; YES), the input operation profile information generation unit 104 uses a dimension reduction method such as a principal component analysis method based on the accumulated input operation information, A one-dimensional evaluation axis that maximizes the average information amount (entropy) of the set of input operation information is calculated (step S18). The calculated evaluation axis is used as the input operation profile information of the respondent, and is stored in the input operation profile information storage unit 105 (step S19). After storing the input operation profile information, the counter Ta is set to 0 (step S20), and the same processing is repeated from step S11.

  FIG. 6 is a flowchart illustrating an example of an operation for calculating a rating value of the subjective rating value detection apparatus 10 according to the first embodiment. The rating value calculation unit 106 inputs the input operation information of interest among the input operation information accumulated in the input operation information storage unit 103 (step S21). The input operation information of interest may be the latest input operation information extracted or may be one selected from past input operation information.

  The rating value calculation unit 106 calls the input operation profile information from the input profile information storage unit 105 (step S22). A plurality of input operation profile information may be stored as time passes. In that case, input operation profile information corresponding to the time of the input operation information of interest is called. Then, using the evaluation axis that is the input operation profile information, the vector of each input operation information having the feature quantity as a dimension is projected onto the evaluation axis, and the coordinate value of each vector after the projection is calculated as the evaluation value. (Step S23). Finally, the calculated rating value is stored in the rating value storage unit 107 (step S24).

  As described above, according to the subjective rating value detection apparatus 10 of the first embodiment, the rating value is calculated from the information input by the respondent intuitively according to the subjective feeling, particularly the flash. Can do. The respondent does not need to determine the strict consistency with the previous answer of the respondent himself regarding the degree of the event when inputting the rating value for the flash (subjective feeling).

  According to the present embodiment, it is not necessary for the respondent to reply to the rating value for the flash that occurred during the activity of the person by converting it to a preset stage and range, and the degree of the flash is input to the input device. It is possible to input from the intuitive operation, and it is possible to automatically calculate the flash value of the respondent's flash from the input operation information.

(Embodiment 2)
FIG. 7 is a block diagram showing a configuration example of the subjective rating value detection apparatus 10 according to Embodiment 2 of the present invention. The subjective rating value detection apparatus 10 according to the second embodiment is based on an answer input terminal device 100 in which a respondent inputs a sense level by operation and records input operation information, and input operation information received from the answer input terminal device 100. A rating value calculation apparatus 200 that calculates a rating value is connected to the network 30 and configured.

  The answer input terminal device 100 includes a switch 101, a feature amount extraction unit 102, an input operation information storage unit 103, and a transmission unit 108. The rating value calculation apparatus 200 includes a receiving unit 201, an input operation profile information generation unit 204, an input operation profile information storage unit 205, a rating value calculation unit 206, and a rating value storage unit 207.

  Also in the second embodiment, the flash will be described as a predetermined feeling that the respondent answers. The switch 101 is operated so that the respondent intuitively inputs the degree of flashing. The feature amount extraction unit 102 extracts input operation information that is a set of predetermined feature amounts from a signal generated by operating the switch 101. Further, time information may be associated with the input operation information. The input operation information storage unit 103 sequentially stores (accumulates) input operation information associated with time information. The transmission unit 108 transmits the accumulated input operation information to the rating value calculation apparatus 200 via the network 30.

  The receiving unit 201 of the rating value calculation apparatus 200 receives input operation information from the answer input terminal apparatus 100 via the network 30. The input operation profile information generation unit 204 generates the input operation profile information of the respondent based on the received input operation information. The input operation profile information storage unit 205 stores the generated input operation profile information.

  The rating value calculation unit 206 calculates the rating value of each input operation information using the input operation profile information stored in the input operation profile information storage unit 205. The rating value storage unit 207 stores the calculated rating value.

  Note that a plurality of answer input terminal devices 100 can be connected to the rating value calculation device 200. In that case, the rating value calculation apparatus 200 generates input operation profile information for each respondent (answer input terminal apparatus 100). For each respondent (answer input terminal device 100), the rating value of the input operation is calculated using the respective input operation profile information.

  FIG. 8 is a flowchart illustrating an example of the operation of the answer input terminal device 100 according to the second embodiment. In the operation of the flowchart of FIG. 8, the operation from the switch operation (step S31) to the input operation information storage (step S35) is the same as the steps S11 to S15 of FIG.

  The answer input terminal device 100 waits for an input operation of the switch 101 to be performed (step S31, step S32; NO). When the respondent recognizes the flashing and intuitively operates the switch 101 with respect to the flashing level (step S32; YES), the feature amount extraction unit 102 receives the input operation signal, and the predetermined feature amount is obtained from the input operation signal. Are extracted (input operation information) (step S33). At the same time, the time of the input operation is counted and associated with the input operation information (step S34). The feature quantity extraction unit 102 accumulates the extracted feature quantity sets (input operation information) in the input operation information storage unit 103 (step S35).

  If not connected to the network 30 (step S36; NO), the process returns to step S31 and is repeated from the switch operation input. If connected to the network 30 (step S36; YES), the transmission unit 108 transmits the untransmitted input operation information stored so far to the rating value calculation apparatus 200 (step S37). After the input operation information is transmitted, the same processing is repeated from step S31.

  FIG. 9 is a flowchart showing an example of the operation of generating input operation profile information in the second embodiment. The rating value calculation apparatus 200 waits for input operation information to be transmitted from the answer input terminal apparatus 100 (step S41; NO). When the input operation information is received by the reception unit 201 (step S41; YES), the input operation profile information generation unit 204 acquires a set of input operation information from the answer input terminal device 100 (step S42). The input operation profile information generation unit 204 generates input operation profile information from the input operation information (step S43) and stores it in the input operation profile information storage unit 205 (step S44).

  The operation of calculating and storing the rating value of the input operation information using the generated input operation profile information is the same as the operation of the first embodiment (FIG. 6).

  As described above, in the second embodiment of the present invention, the input operation profile device that analyzes input operation information and the rating value calculation device 200 are configured to be separated from the answer input terminal device 100 of the respondent. Therefore, it is possible to reduce the burden of operation by downsizing the respondent's terminal device.

  FIG. 10 is a block diagram illustrating a physical configuration example of the subjective rating value detection apparatus 10 according to the embodiment of the present invention. In the case of the subjective rating value detection device 10 according to the second embodiment, each of the answer input terminal device 100 and the rating value calculation device 200 is configured as shown in FIG.

  As shown in FIG. 10, the subjective rating value detection apparatus 10 includes a control unit 21, a main storage unit 22, an external storage unit 23, an operation unit 24, a display unit 25, and a transmission / reception unit 26. In the case of the first embodiment, the transmission / reception unit 26 may not be provided. The main storage unit 22, the external storage unit 23, the operation unit 24, the display unit 25, and the transmission / reception unit 26 are all connected to the control unit 21 via the internal bus 20.

  The control unit 21 is composed of a CPU (Central Processing Unit) and the like, and in accordance with the control program 29 stored in the external storage unit 23, the above-described subjective rating value detection process, response input transmission or input operation profile generation, and rating value calculation Execute the process.

  The main storage unit 22 is composed of a RAM (Random-Access Memory) or the like, loads a control program 29 stored in the external storage unit 23, and is used as a work area of the control unit 21.

  The external storage unit 23 includes a nonvolatile memory such as a flash memory, a hard disk, a DVD-RAM (Digital Versatile Disc Random-Access Memory), a DVD-RW (Digital Versatile Disc ReWritable), and the processing described above is performed by the control unit 21. A control program 29 to be executed is stored in advance, and data stored in the control program 29 is supplied to the control unit 21 in accordance with an instruction from the control unit 21, and the data supplied from the control unit 21 is stored.

  The operation unit 24 includes a pointing device such as a keyboard and a mouse, and an interface device that connects the keyboard and the pointing device to an internal bus. The operation of the respondent is input via the operation unit 24 and supplied to the control unit 21.

  The display unit 25 includes a CRT (Cathode Ray Tube) or an LCD (Liquid Crystal Display), and displays input operation information, input operation profile information, or a rating value.

  The transmission / reception unit 26 includes a wireless transmitter / receiver, a wireless modem or a network termination device, and a serial interface or a LAN (Local Area Network) interface connected thereto. Input operation information is transmitted / received via the transmitter / receiver 26. That is, the answer input terminal device 100 transmits input operation information via the transmission / reception unit 26, and the rating value calculation device 200 receives input operation information via the transmission / reception unit 26. In the case of the subjective rating value detection apparatus 10 according to the first embodiment, the input operation information and the rating value can be transmitted to a server (not shown) via the transmission / reception unit 26.

  The processing of the feature amount extraction unit 102, the input operation profile information generation units 104 and 204, the rating value calculation units 106 and 206, the transmission unit 108, and the reception unit 201 of the subjective rating value detection apparatus 10 is performed by the control program 29 and the control unit 21. The main storage unit 22, the external storage unit 23, the operation unit 24, the display unit 25, the transmission / reception unit 26, and the like are used as resources for processing.

  Other suitable modifications of the present invention include the following configurations.

About the subjective rating value detection apparatus according to the first aspect of the present invention,
Preferably, the profile generation means is a one-dimensional unit that maximizes an average information amount (entropy) of a set of values extracted for two or more predetermined feature amounts of the input operation using a predetermined dimensional reduction method. Calculate the evaluation axis.

  Preferably, the predetermined feature amount of the input operation includes the number of times the switch is operated at intervals of a predetermined time or less, the frequency of operating the switch per predetermined time, the time during which the switch is operated, The interval of operating the switch, the magnitude of the operation of the switch, the speed of operating the switch, the magnitude of the force operating the switch, the magnitude of the moment operating the switch, or a combination thereof It is characterized by.

  Preferably, the predetermined sensation represents a flashing level of a user's thought.

About the subjective rating value detection method according to Embodiment 2 of the present invention,
Preferably, in the profile generation step, a one-dimensional value that maximizes an average information amount (entropy) of a set of values extracted for two or more predetermined feature amounts of the input operation using a predetermined dimensional reduction method. Calculate the evaluation axis.

  Preferably, in the feature extraction step, the number of times the switch is operated at an interval equal to or less than a predetermined time, the frequency of operating the switch per predetermined time, the time during which the switch is operated, and the interval at which the switch is operated A combination including any two or more of the magnitude of the operation of the switch, the speed at which the switch is operated, the magnitude of the force to operate the switch, or the magnitude of the moment to operate the switch The predetermined feature amount is extracted from the input operation information.

  Preferably, the predetermined sensation represents a flashing level of a user's thought.

  In addition, the above-described hardware configuration and flowchart are examples, and can be arbitrarily changed and modified.

  A central part that performs device control processing including the control unit 21, the main storage unit 22, the external storage unit 23, the operation unit 24, the internal bus 20, and the like uses a normal computer system, not a dedicated system. Is feasible. For example, a computer program for executing the above operation is stored and distributed in a computer-readable recording medium (flexible disk, CD-ROM, DVD-ROM, etc.), and the computer program is installed in the computer. Thus, the subjective rating value detection apparatus 10 that executes the above-described processing may be configured. Alternatively, the subjective rating value detection apparatus 10 may be configured by storing the computer program in a storage device included in a server device on a communication network such as the Internet, and downloading it by a normal computer system.

  Further, when the function of the subjective rating value detection device 10 is realized by sharing an OS (operating system) and an application program, or by cooperation between the OS and the application program, only the application program portion is recorded on a recording medium or a storage device. May be stored.

  It is also possible to superimpose a computer program on a carrier wave and distribute it via a communication network. For example, the computer program may be posted on a bulletin board (BBS, Bulletin Board System) on a communication network, and the computer program distributed via the network. The computer program may be started and executed in the same manner as other application programs under the control of the OS, so that the above-described processing may be executed.

  INDUSTRIAL APPLICABILITY According to the present invention, the present invention can be applied to uses such as an information collection device that continuously collects personal flash rating values in a long time and in real time, and a program for realizing the information collection device on a computer. Further, the present invention can be applied to an application such as a reply terminal for subjective information evaluation experiments in which a change in operation that appears unconsciously in the input of the respondent is reflected in the rating value.

DESCRIPTION OF SYMBOLS 10 Subjective rating value detection apparatus 21 Control part 22 Main memory part 23 External storage part 24 Operation part 25 Display part 26 Transmission / reception part 29 Control program 100 Answer input terminal device 101 Switch 102 Feature-value extraction part 103 Input operation information storage part 104 Input operation Profile information generation unit 105 Input operation profile information storage unit 106 Rating value calculation unit 107 Rating value storage unit 108 Transmission unit 200 Rating value calculation device 201 Reception unit 204 Input operation profile information generation unit 205 Input operation profile information storage unit 206 Rating value calculation Part 207 Rating value storage part

Claims (9)

A switch that allows a user to intuitively input a predetermined sense according to its subjectivity,
Feature extraction means for extracting a set of the predetermined feature values from the input operation of the switch;
Storage means for storing the set of feature quantities extracted by the feature extraction means;
Profile generation means for obtaining an evaluation axis that maximizes the average information amount (entropy) of the predetermined feature quantity set of the input operation of the user based on the feature quantity set accumulated in the storage means;
Rating means for calculating a rating value obtained by projecting the set of predetermined feature values extracted from one of the input operations onto the evaluation axis obtained by the profile generating means;
A subjective rating value detection apparatus comprising:   The profile generation means uses a one-dimensional evaluation axis that maximizes an average information amount (entropy) of a set of values extracted for two or more predetermined feature amounts of the input operation using a predetermined dimension reduction method. The subjective rating value detection apparatus according to claim 1, wherein the subjective rating value detection apparatus calculates the subjective rating value.   The predetermined feature amount of the input operation is the number of times the switch is operated at intervals of a predetermined time or less, the frequency of operating the switch per predetermined time, the time during which the switch is operated, and the switch operation. It includes an interval, a magnitude of operation of the switch, a speed of operating the switch, a magnitude of force for operating the switch, a magnitude of moment for operating the switch, or a combination thereof. The subjective rating value detection apparatus according to claim 1 or 2.   The subjective rating value detection apparatus according to claim 1, wherein the predetermined feeling represents a flashing level of a user's thought. A step of accepting an input for the user to operate the switch intuitively according to the subjectivity for a predetermined sense;
A feature extraction step of extracting the predetermined set of feature values from the input operation of the switch;
A storage step of storing the set of feature quantities extracted in the feature extraction step;
A profile generation step for obtaining an evaluation axis that maximizes an average information amount (entropy) of a predetermined feature amount set of the input operation of the user based on the feature amount set accumulated in the storage step;
A rating step for calculating a rating value obtained by projecting the set of predetermined feature values extracted from one of the input operations onto the evaluation axis obtained in the profile generation step;
A subjective rating value detection method characterized by comprising:   In the profile generation step, a one-dimensional evaluation axis that maximizes an average information amount (entropy) of a set of values extracted with respect to two or more predetermined feature amounts of the input operation using a predetermined dimensional reduction method. 6. The subjective rating value detection method according to claim 5, wherein the subjective rating value is calculated.   The feature extraction step includes the number of times the switch is operated at intervals of a predetermined time or less, the frequency of operating the switch per predetermined time, the time during which the switch is operated, the interval at which the switch is operated, the switch A combination including any two or more of the magnitude of the operation, the speed at which the switch is operated, the magnitude of the force to operate the switch, or the magnitude of the moment to operate the switch The subjective rating value detection method according to claim 5 or 6, wherein a quantity is extracted from the input operation information.   The subjective rating value detection method according to claim 5, wherein the predetermined feeling represents a flashing level of a user's thought. On the computer,
A step of accepting an input for the user to operate the switch intuitively according to the subjectivity for a predetermined sense;
A feature extraction step of extracting the predetermined set of feature values from the input operation of the switch;
A storage step of storing the set of feature quantities extracted in the feature extraction step;
A profile generation step for obtaining an evaluation axis that maximizes an average information amount (entropy) of a predetermined feature amount set of the input operation of the user based on the feature amount set accumulated in the storage step;
A rating step for calculating a rating value obtained by projecting the set of predetermined feature values extracted from one of the input operations onto the evaluation axis obtained in the profile generation step;
A program characterized by having executed.

Images