JP2007052601A - Usability-evaluating device, usability-evaluating method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To particularly evaluate how easily a user learns and gets accustomed to functions of equipment when the usability of the equipment is evaluated. <P>SOLUTION: A usability-evaluating device which is provided with an operation inputting part 12 receiving a user's request, and a biological signal detecting part 15 measures event-related electric potential which can be measured by an electroencephalograph as a user's reaction posterior to the operation of equipment. A function-evaluating part 16 evaluates how easily a user learns each function, how much a user is interested in each function and how deeply a user is skilled in each function according to the amplitude and a latent time of the event-related electric potential which periodically changes in response to the inner state of a user. By this, for example, it can be determined which function can easily be learned among the functions which are determined not to be operated as an action evaluation, and then a more detailed usability evaluation becomes possible. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a usability evaluation apparatus, a usability evaluation method, and a program for evaluating usability of a device using a user's biological signal.

  Various devices used by users need to be evaluated in advance for their ease of use. Ease of use of a user's device is called usability, and by appropriately performing this usability evaluation, the evaluation result can be used for product design correction and next device specification determination.

  As conventional usability evaluation methods, there are (1) a method mainly collecting user's subjective evaluation in a questionnaire format, (2) a method based on behavior observation such as user operation history information, and (3) a method using physiological signals. . The following is an overview of conventional technology.

  Subjective evaluation is based on questionnaires and interviews by listening to subjective impressions regarding ease of use. Subjective evaluation is said to be a part of feeling that evaluation results are influenced by various factors and that it is possible to respond subjectively.

  The behavior evaluation is performed by observing the user's facial expression and behavior from the user's operation history information and the video of the user and analyzing the data. For example, the user is in an emotional state or perplexed. Evaluated in connection with the state. In addition, the behavior evaluation focuses on the operation while the user operates the device, and there are many things that can be measured by external observation such as the operation content and operation time interval. By using the user's action record, the achievement degree of the user's device operation task, the user's degree of confusion, the user's impression of the device operation, and the like can be evaluated.

  As an example of behavior evaluation, in the active help device of Patent Document 1, in the interactive operation using the screen, the stop time of each display screen is measured and compared with the standard display time set for each display screen. A method for evaluating whether or not the above-described delay or sticking has occurred in the interaction between the user and the system is used, and the evaluation result is used to present help information.

  Next, physiological evaluation is to acquire physiological signals related to the user's behavior and state mainly by a sensor worn on the user and use it for evaluation. Typical physiological signals include heart rate, pulse, blood pressure, brain waves, etc., and the characteristic is that, unlike subjective responses such as questionnaires, it is possible to record changes in physiological state that the user himself is not aware of In addition, it is possible to record a state that cannot be known by observation from the outside.

  An example of physiological evaluation is described in Patent Document 2. In the worker's psychosomatic state evaluation method of Patent Document 2, a blink sensor and skin impedance are used as a biosensor to evaluate the worker's mental and physical state, and physiological information is analyzed by a physiological information analysis unit to generate stress. Whether it is present or relaxed is determined. Moreover, in patent document 2, it is supposed that more appropriate evaluation will be attained by combining an action index with these physiological information.

Thus, usability evaluation has been performed using behavioral indices and physiological indices that can be acquired from the user.
JP 2000-227826 A JP-A-11-65422 WO2005 / 001677 ISO 9241-11 (3. Definition (1998)) Kaga et al., “Event-related Potential (ERP) Manual-Focusing on P300”, Shinohara Publishing Shinsha, 1995 Atsuko Harada “What is“ Ease of Use ”as Seen from the Elderly: From a Cognitive Engineering Perspective”, Japan MEE Society Special Study Group, Human Interface Study Group in the Medical and Welfare Field, 2004 Second Proceedings of the symposium,

  However, there is a point that cannot be measured even by the above-described conventional technology. That is, from the viewpoint of a slight change of the user state in a short time, how easy the user can learn the function, and how easy it is to get used to.

  Here, the definition of usability will be described. Usability is defined in the international standard ISO 9241-11 (Non-Patent Document 1). There are: “Effectiveness, Efficiency, and Satisfaction when a product is used by a particular user to achieve a specified goal in a particular usage situation. ) Degree ".

  Among these, “effectiveness” indicates the accuracy and completeness in achieving the specified goal by the user, and is an index of how much the operation has been achieved without performing an incorrect operation.

  “Efficiency” is a resource that the user spends on accuracy and completeness to achieve the goal, and is an indicator of how much effort the user has had to achieve the correct operation. . The hardships here include operational stress and the possibility of making mistakes without making mistakes.

  The “satisfaction level” is a sense of discomfort and a positive attitude when using the product. This arises from the impression and comfort of the product.

  In usability evaluation, it is necessary to comprehensively evaluate from various viewpoints as described above. According to this definition, what can be measured in Patent Documents 1 and 2 is “effectiveness” regarding accuracy in the above definition, “efficiency” in terms of operational stress, and “satisfaction” in the sense of discomfort. The corresponding item.

  That is, depending on the method using the behavior index exemplified in Patent Document 1, the mental state at the time of the user's operation that does not appear in the behavior has not been understood.

  In addition, depending on the method using the physiological index and the method combining the physiological index and the behavioral index as shown in Patent Document 2, the state of the user such as the degree of tension, the degree of concentration, and the degree of fatigue is mainly determined by the physiological index. An evaluation was made. However, in this way of thinking, in relation to the overall evaluation of the user's device, even though it can be evaluated that the device can be used with less stress when it is used, or it can be used relaxedly, There was no evaluation for ease of remembering or discomfort in operation.

  By the way, the user does not care only about whether or not the operation of the device is possible. For example, on the product evaluation page sent by the user on the Internet, about the operability of the equipment, “If you operate for 2 hours repeatedly, you will get used to it, but you may still feel a little operation,” A description such as “I think that the possibility of erroneous operation is quite low, but I feel that it is difficult to operate intuitively” can be seen. As for the former case, since the user always learns even when operating the device, the operation is realized by the adaptability of the user, but it is suggested that even if it can be realized, the operation may be uncomfortable. The latter case also suggests that there is a case where the operation is not mistaken and that the intuitive operation can be independent.

  These examples include ease-of-use in the definition of usability, “efficiency” in the sense that mistakes are not repeated, and non-intuitive operation methods that can be operated only by remembering the operation of the device when performing operations. Related to “satisfaction” in the sense that In other words, it is related to how much trouble the user is forced to learn to operate the device and how easy it is to learn and become familiar with the operation of the device.

  With regard to these points, it has been said that the user should be able to achieve the operation in the behavioral evaluation of whether or not the operation can be performed conventionally. In addition, there are many reasons why the user has a gap in the time interval from when an instruction is given until the operation is executed, which is considered to be relatively related to proficiency in behavior evaluation. Was unknown.

  As described above, the “efficiency” regarding ease of remembering has been confirmed only by subjective reflection reports so far.

  With respect to such a problem, the inventors of the present application focused on event-related potentials that can be measured with an electroencephalograph as shown in Non-Patent Document 2. Inventors have already devised a service providing apparatus using a signal that appears characteristically when the user's expectation of device operation differs from the actual device operation in Patent Document 3. . Japanese Patent Application Laid-Open No. 2004-151561 shows that the operation of the device is as intended by the user or can be determined by the signal but not as intended by the user.

  Further, from the viewpoint of usability evaluation, if the method shown in Patent Document 3 is used, it is possible to evaluate which function the user has not been able to operate as intended among a plurality of functions of the device.

  Here, an example where the user's expectation for the device operation is different from the actual device operation will be briefly described with reference to FIG. FIG. 2A shows a scene in which the user 101 operates the television 103 using the remote controller 102.

  In this scene, the flow as shown in FIG. Assume that the user 101 wants to watch the television 103 in step S14. In accordance with the desire, the user 101 operates the remote controller 102 in step S15 to try to realize watching TV. Here, although the user 101 may not be explicitly aware, it is considered that the television 103 is projected by the operation of the remote controller 102.

  In the next step S16, when the television is displayed as usual, in step S17, the television can be watched as usual, but if for some reason (the failure of the TV 103 or the remote control 102, the wrong button on the remote control 102, etc.) If the television 103 does not operate as expected by the user 101, the television 103 is not shown in step S18. In this case, the user 101 thinks “That?

  The matter described in Patent Document 3 deals with the fact that the situation of step S19 can be clearly distinguished from the case of step S17 when the event-related potential is used. In the present invention, the signal confirmed in this case is hereinafter referred to as a disappointment signal.

  In view of the above-mentioned problems of the prior art, the present invention detects a user's mental state using a biological signal in the usability evaluation of a device, and finally learns about a specific function of the device, The purpose is to perform a more detailed usability evaluation by evaluating whether the function is easy to learn or easy to use.

  In order to solve the conventional problem, the usability evaluation apparatus of the present invention includes an operation input unit that receives a user request, and a response content determination unit that determines a response content with respect to the request received by the operation input unit. A presentation unit that presents the response content determined by the response content determination unit to the user, a biological signal detection unit that measures an event-related potential of an electroencephalogram as the biological signal of the user, and the measured event-related A storage unit that stores potential data, and the event-related potential measured by the biological signal detection unit starting from the timing at which the presentation unit presents the response content, after a predetermined time from the start point A specific event-related potential determination unit that determines whether or not there is a specific event-related potential, and stores the determination result in the storage unit, and the storage unit stores the specific event-related potential determination unit. In the data of a plurality of the event-related potential with, in which and a user evaluation unit which evaluates the state of the user from the degree of time-series changes of a plurality of the specific event-related potentials. Here, the degree of change in time series of a plurality of specific event-related potentials refers to the specific event-related potentials as the detection order increases when comparing specific event-related potentials detected multiple times in time series. Is the degree to which changes. In addition, the presence or absence of the specific event-related potential is determined from data in a predetermined detection target section represented by time of the event-related potential.

  By having such a configuration, when a specific event-related potential such as a “disappointment signal” is detected multiple times, it is observed that this specific event-related potential changes in time series. The user state is evaluated based on the degree of change, and the usability defined in Non-Patent Document 1 is evaluated over a wide range.

  In a preferred embodiment, the user evaluation unit evaluates a user state based on a time-series change in a peak amplitude of the waveform of the specific event-related potential. The waveform of the specific event-related potential is a waveform represented by the data in the predetermined detection target section, and the peak amplitude is a large peak-shaped waveform having a positive potential vertex of the waveform at the vertex. It is the magnitude of the potential.

  In a preferred embodiment, the user evaluation unit evaluates a user state based on a time-series change in a peak latency of a waveform of the specific event-related potential. Here, the peak latency is an elapsed time from the starting point of the peak of a large mountain-shaped waveform having a positive potential vertex among the waveforms represented by the data in the predetermined detection target section.

  In a preferred embodiment, the user evaluation unit evaluates a user state based on a time-series change in a waveform shape of the specific event-related potential.

  In a preferred embodiment, the user evaluation unit evaluates, as a user state, a degree to which the user learns an operation of the function when the user operates a specific function of a predetermined device a plurality of times. The ease of remembering the operation of the function is determined.

  The user evaluation unit compares each degree of time-series change of the specific event-related potential in the operation of two functions, and the function operation is performed when the amplitude of the waveform of the specific event-related potential decreases earlier. It is preferable to further determine that it is easy to remember.

  In a preferred embodiment, the user evaluation unit evaluates the degree of interest of the user with respect to a specific function provided by a predetermined device as a user state.

  The user evaluation unit compares the degree of time-series change of the specific event-related potential when the two specific functions are provided, and the degree of decrease in the amplitude of the waveform of the specific event-related potential It is preferable to determine that the interest is sustained when there is less.

  In a preferred embodiment, the user evaluation unit evaluates a user's proficiency with respect to an operation of a specific function in a predetermined device as a user state.

  It is preferable that the user evaluation unit determines that the user has mastered the operation of a specific function when the amplitude of the waveform of the specific event-related potential decreases.

  The predetermined time is preferably 450 to 800 msec.

  The predetermined time is preferably around 600 msec or around 750 msec.

  The usability evaluation method of the present invention includes a step of receiving a user request, a step of determining a response content in response to the request, a step of presenting the response content to the user, and an electroencephalogram event as the user's biological signal From the step of measuring the related potential, the step of accumulating data of the measured event-related potential, and the timing of presenting the response content, from the measured event-related potential, from the starting point Determining whether or not there is a specific event-related potential after a predetermined time, and accumulating the determination result, and a plurality of the event-related potentials having the specific event-related potential, Evaluating the state of the user from the degree of time-series change of the specific event-related potential.

  The program of the present invention includes a step of receiving a user's request, a step of determining a response content in response to the request, a step of presenting the response content to the user, and an event-related potential of an electroencephalogram as the user's biological signal A step of accumulating data of the measured event-related potential, and a timing at which the response content is presented as a starting point, from the measured event-related potential, a predetermined value from the starting point Determining whether or not there is a specific event-related potential after time, and accumulating the determination result, and a plurality of the specific events in the data in which the plurality of event-related potentials having the specific event-related potential are stored A program for causing a computer to execute the step of evaluating the state of the user from the degree of time-series change of the related potential. That.

  According to the usability evaluation apparatus of the present invention, it is possible to evaluate the ease of remembering the function, the degree of interest, the proficiency level, etc., which were difficult to measure in the past.

  This isn't about whether or not you can remember, but you can evaluate how much you can remember but how much it costs to remember, and you can use the results to design and improve a device that is easy to use.

  In addition to what has been described in the section of the problem to be solved by the invention, the present inventors changed both the amplitude and the latency of the disappointment signal when the disappointment situation occurred multiple times. Newly discovered that it is related to the degree of disappointment. Details of these experiments will be described in detail in the embodiment. In the present invention, attention is paid to the fact that the change in the degree of disappointment can be applied to the evaluation of usability “efficiency”.

  The usability evaluation apparatus of the present invention is determined by an operation input unit that receives a user request, a response content determination unit that determines a response content with respect to the request received by the operation input unit, and the response content determination unit Starting from a presentation unit that presents the response content to the user, a biological signal detection unit that measures an event-related potential of an electroencephalogram as the user's biological signal, and a timing at which the presentation unit presents the response content. Among the event-related potentials measured by the signal detection unit, a specific event-related potential determination unit that determines the presence or absence of a specific event-related potential after a predetermined time from the starting point, and the measured event-related potential An accumulation unit for accumulating data and a determination result of the presence or absence of the specific event-related potential; and the data in which a plurality of the event-related potentials having the specific event-related potential are accumulated In this case, since it comprises a user evaluation unit that evaluates the state of the user from the degree of time-series change of the specific event-related potential, when the device reacts according to the operation of the user, A change in the strength of the “disappointment” that the user has can be detected by an event-related potential and used for usability evaluation. This change in disappointment is considered to be related to how familiar the user is with the operation of the device and the friendliness of the operation of the device. This evaluation is not an evaluation of whether or not it was used or stressed, which was evaluated by the conventional evaluation method, but it can be evaluated how much a particular function is easy to learn, how easy it is to use, etc. Qualitatively new items can be added to evaluation measurement items. Here, the specific event-related potential is, for example, potential data within a predetermined time range of a certain kind of event-related potential that appears after a predetermined time from the starting point when a “disappointment” occurs. You can also.

  Evaluation of the ease of remembering of these functions and devices has only relied on subjective evaluation, and the present invention enables quantitative evaluation for the first time.

  The disappointment signal, which is a kind of specific event-related potential, is a large mountain-shaped waveform having a positive potential peak at about 300 msec to 600 msec starting from the timing when the presentation unit presents the response content. Is about 300 msec to the amplitude (the magnitude of the potential at the time when the peak of the positive potential is taken) and the latency (the time from the stimulus presentation at the time when the peak of the positive potential is taken). Often seen around 600 msec or 750 msec).

  The user state can be evaluated by observing the degree of time-series change in amplitude, which is a feature of the disappointment signal (specific event-related potential).

  The evaluation can also be performed by looking at the degree of time-series change in latency, which is another feature other than amplitude.

  In addition, since the waveform of the specific event-related potential varies depending on the experimental conditions, it is possible to evaluate not only the vertex information such as latency and amplitude but also the overall waveform shape. As a waveform evaluation, there is a method of calculating a distance using the entire waveform as a template.

  More specifically, the above-described user state evaluation makes it possible to evaluate the ease of device operation remembering for a specific function of the device. This makes it possible to evaluate the ease of remembering each function of the device, rather than evaluating the ease of use of the entire device as in the past. As an evaluation method, it is conceivable that the function is easier to remember if the amplitude in the detection target section decreases earlier.

  By such processing, more detailed device evaluation is realized.

  The user evaluation unit can also evaluate the degree of interest in a specific function provided by the device. An information recommendation function or the like can be considered as a function provided by the device.

  Here, it can be judged that the interest is maintained if the amplitude of the disappointing signal does not decrease, and that the information recommendation software is good. This makes it possible to evaluate the ease of getting tired, which is the degree of interest in the information recommendation content, which has been difficult to evaluate in the past.

  The user evaluation unit can also evaluate the proficiency level of a specific user for a specific function in device operation. For example, if the amplitude in the detection target section of the disappointment signal decreases, it can be evaluated that the user has mastered a specific function.

  Even in a complicated information device having different proficiency levels for each individual by this evaluation, the proficiency level for each individual can be evaluated.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the following drawings, components having substantially the same function are denoted by the same reference numerals for the sake of brevity.

(Embodiment 1)
When a user uses a device, it may be possible to use various functions of the device at various frequencies. In terms of ease of use of the device, it is considered that the ease of use of the entire device is determined by the total evaluation value of how easy each function is to use.

  First, a processing method in the case where the usability evaluation apparatus according to Embodiment 1 is applied to the evaluation of the ease of remembering the device operation will be described. This assumes the case where the evaluation target device is evaluated in a room for an evaluation experiment. Evaluation target devices are devices that have one or more functions that can be operated by the user. For example, audiovisual products such as TVs, videos, and DVDs, home appliances such as microwave ovens, rice cookers, and washing machines, personal computers, This includes information devices such as car navigation systems and mobile phones. In the following example, a description will be given mainly assuming an operation of a television or a DVD.

  FIG. 1 shows an example of the configuration of a usability evaluation apparatus in the present embodiment. 10 is a user, 11 is an operation instruction unit, 12 is an operation input unit, 13 is a device reaction unit (presentation unit), 15 is a biological signal detection unit, 16 is a function evaluation unit (user evaluation unit), 17 is a storage unit, 18 Is a specific event-related potential determination unit, and 19 is a response content determination unit. The operation input unit 12, the device reaction unit 13, and the response content determination unit 19 are collectively referred to as an evaluation target device 14. The evaluation target device 14 may be an independent device or may be implemented in a form incorporated in the evaluation device.

  In the usability evaluation, the user 10 is sequentially presented with operation instructions set in advance by the evaluator from the operation instruction unit 11. Note that the operation instruction can be given verbally or by a document from the evaluator. In this case, the operation instruction unit 11 is not necessary in the usability device.

  Based on the operation instruction, the user 10 performs an operation that the user 10 considers appropriate (a request from the user 10) via the operation input unit 12. The operation input unit 12 is, for example, a mouse or a keyboard for computer evaluation, or a remote control panel for DVD or television evaluation.

  In response to the operation of the user 10, the evaluation target device 14 determines the device operation (response content) in the response content determination unit 19 according to the operation of the user 10, and the operation is displayed in the device reaction unit 13. Examples of the operation include changes in characters and drawings displayed on the screen, changes in sound, and the like.

  Next, the biological signal detection unit 15 measures a biological reaction of the user 10 when viewing the device operation by the device reaction unit 13, particularly a signal centered on an event-related potential measurable with an electroencephalograph. Signals measured by the biological signal detection unit 15 (data in which measurement time and potential are set as one set) are stored in the storage unit 17. The accumulated data is determined by the specific event-related potential determination unit 18 for the presence or absence of the specific event-related potential. The presence / absence of the specific event-related potential is also stored in the storage unit 17. Here, the specific event-related potential is a specific event-related potential after a predetermined time from the timing when the device reaction unit 13 presents the device operation to the user, and an example is a disappointing signal. it can. The function evaluation unit 16 performs function evaluation based on the data stored in the storage unit 17 and the current data sent from the biological signal detection unit 15.

  The detailed operation will be described below with reference to the configuration diagram of FIG. 1 and the flowchart of FIG.

  FIG. 7 is a flowchart outlining the entire process of the usability evaluation apparatus. Broadly divided into four parts. First, the overall flow will be described.

  Although there are three embodiments described in this specification, including this embodiment, steps S20 to S60 can be commonly used in all the embodiments in the flowchart of FIG. Yes, the user evaluation method in step S80 differs depending on each embodiment.

  The first step is an experiment execution and data collection step (step S20). Here, the user performs various device operations on the evaluation target device, the device operates by the operation, and a biological signal generated by the user by the operation is measured and recorded. The biological signal includes, for example, an event-related potential that can be measured with an electroencephalograph, and a waveform that is a temporal change in the event-related potential is recorded. The measurement may be performed over the operation of a plurality of functions, or may be performed in the operation of a plurality of subjects to record a waveform.

  The next step is a step of processing the recorded event-related potential (step S40). Here, from the recorded event-related potentials, it is calculated by processing the waveform of each event-related potential whether or not a disappointing signal that is a specific event-related potential of interest in the present invention is included. As a result, the signal characteristics are calculated at the same time whether or not the unexpected signal is included or not. Examples of the characteristics of the signal include the amplitude and latency of a positive peak potential when a disappointment signal is included, and these can be used as an index for evaluating the user state.

  Next, after the disappointment signal detection process is completed, a waveform addition process is performed (step S60). Here, the noise of the electroencephalogram signal can be reduced by adding a plurality of disappointing signals after selection based on the presence or absence of the disappointing signals.

  This process is used when it is not possible to determine a decrease in waveform unless addition is performed, or when it is necessary to determine a decrease in waveform over a large number of times. This process is not always necessary when there is sufficient discrimination performance even with non-addition waveforms.

  The last step is a step of performing user evaluation based on the processed signal (step S80). This step makes it possible to evaluate how easy the operation was to learn in each function.

  In the following, three types of embodiments will be described. Basically, among the above steps, Steps S20 to S60 can be processed in a substantially common way of thinking. It is particularly different.

  Before going into the details of the processing of each rough step below, we will explain the outline of the disappointment signal acquisition experiment, that the specific event-related potential is a signal that is uniquely seen in the disappointment situation, and that the amplitude of the disappointment signal And latency changes each time a subject encounters a similar situation, indicating that the user has to do with adapting to the situation.

  First, an experiment conducted by the present inventors for obtaining a disappointing signal will be described.

  Fig.3 (a) is a figure which shows the outline | summary of an experimental procedure. In this experiment, an instruction is given to the subject (Procedure A), and the subject operates the device thinking of necessary actions (Procedure B) in response to the instruction (Procedure C). ).

  First of all, the experiment operator gives the subject a “L” or “R” character displayed on the screen, so the left click of the mouse for the L character and the right for the R character. I want you to press the click. ” Then, “L” and “R” are randomly selected with a probability of 50% and displayed on the screen (procedure A). The subject looks at the displayed characters and clicks right or left according to the taught rule (procedure B). Then, whether or not the operation is correctly clicked is displayed on the screen by “◯” and “×” (procedure C).

  However, in this experiment, even if the click is made correctly (should be done almost 100% correctly), “x” is displayed with a probability of 20%. From the subject's point of view, if they clicked correctly, they expected that “O” would be displayed, but if “X” was displayed, they would think “That”. In other words, the situation is “disappointing”, which was not exactly what was expected. The purpose of this experiment is to confirm whether this “disappointing” situation cannot be detected from event-related potentials in the electroencephalogram.

  FIG. 3B is a flowchart showing a procedure for one trial. First, the characters “L” or “R” are selected and displayed on the screen with a probability of 50% (S1101), and the subject determines which button to click by operating the display and operates the mouse. (S1102). In response to the subject's operation, “◯” and “×” are displayed based on whether or not the operation is correct. At this time, even when “O” should be displayed, “X” is displayed with a probability of 20% (S1103). The event-related potential in the subject's electroencephalogram is measured starting from the timing at which “O” and “x” are displayed (S1104), and the measured event-related potential is processed to identify the disappointment signal (S1105). .

  In the experiment, for a plurality of subjects, first, 30 trials in which “◯” is always displayed as practice were performed, and then 100 trials according to the procedure shown in FIG.

  FIG. 4 is a graph showing the result of the experiment, showing a waveform obtained by averaging the experimental data of four subjects. The graph of FIG. 4 is obtained by adding the potential waveform measured by an electroencephalograph to about 20 times each starting from the presentation of “◯” and “×”, and a total of 82 times for four subjects. The time from the stimulus presentation (when “◯” and “×” are displayed) is the unit msec, the vertical axis is the potential, and the unit is μV. The thick line 105 is “disappointment”, that is, the waveform when “x” is displayed even though it is correctly clicked, and the thin line 104 is the waveform when it is normal, that is, when it is correctly clicked and “◯” is displayed.

  From FIG. 4, it can be seen that, when “disappointing”, an event-related potential (specific event-related potential) having characteristics different from the normal time appears around 600 msec after the stimulus presentation. That is, it is expected that the user's “disappointing state” could be detected by measuring this event-related potential. As a result of the t-test, the time when a significant difference was observed in the vicinity of the latency of 450 to 800 msec depending on whether or not there was a disappointment (p = 0.05) is shown in a band 106.

  In addition, according to the international 10-20 method, the electrode was affixed on four places, 1) Pz: midline parietal, 2) 3) A1, A2: binaural fistula, 4) body earth (Z): nose root. The sampling frequency was 1000 Hz, and the band pass filter was 0.03-20 Hz.

  According to another experiment conducted by the present inventors, it has been found that an event-related potential with different latencies appears in the signal of the disappointment state depending on the complexity of the stimulus and the direction of attention to the subject's stimulus. Yes. For example, when a user watching TV switches a channel to a news program, if the channel is switched to a news program of the same category but the channel is switched to a channel different from the news program that the user wants to watch, the disappointing signal is It appears around 750 msec, not around 600 msec. This is different from programs in different categories (for example, news programs and baseball programs or news programs and animation programs) in the same category of news programs. I think that latency will occur.

  In addition, the event-related potential 105 measured in this experiment has a longer response time from the stimulus presentation than P300 shown in Non-Patent Document 2 and the like. The reason is considered as follows.

  In other words, what has been conventionally performed with respect to event-related potentials is an experiment under a passive condition in which differences in pitch or image are judged. On the other hand, in the experiment conducted by the inventors of the present application, the user selects and executes an action with a certain expectation, and determines whether the expectation and the actual device operation are the same. The point that the subject selects the action with expectation is significantly different from the previous one. The determination as to whether or not the operation as expected has been performed is a higher-order determination as compared to the case of merely determining the difference in pitch or image. For this reason, it is considered that the reaction time from the stimulus presentation is later than the measurement so far, and a peak appears at about 600 msec after the stimulus presentation.

  As is clear from the above experiment, there is a clear difference in event-related potentials measured by an electroencephalograph depending on whether or not “◯” and “x” are displayed as expected by the subject. Therefore, this specific event-related potential can be used as a “disappointing signal” in determining the user state.

  Next, a description will be given of what amplitude and latency changes when this disappointment signal is repeated.

  The disappointment signal reflects the brain activity related to “disappointment” when the subject selects an action with expectation and the actual device behavior differs from the expectation. Since the event-related potential has a low S / N, in order to clarify the component of the disappointing signal, FIG. In the case of a component that reflects the mismatch between the subject's expectation and the result, such as a disappointment signal, the component that appears along with the change in expectation may change.

  Specifically, in the experiment of obtaining a disappointment signal, the inventors of the present application may present an “x” even though the subject correctly selects an action while the disappointment trial is repeated. As a result, it was thought that the component of the disappointing signal may change due to the change in the expectation of the subject. Based on this idea, the processing shown in the flowchart of FIG. 5 was performed, and the time series change of the disappointing signal was examined.

  S1301 is a step of sampling the potential change of the electroencephalogram starting from the timing at which the response content is presented (when the response is presented). Examples of the sampling frequency include 200 Hz, 500 Hz, and 1000 Hz.

  S1302 is a step of removing noise from the potential change waveform sampled in S1301. Here, it passes through a bandpass filter of, for example, 0.03 to 30 Hz in order to cut low frequency components and high frequency components mixed in the signal, or has an amplitude of 80 μV or more in order to reduce the influence of blinking or eye movement. Remove the waveform from the identification target.

  S1303 is a step of determining whether or not “x” is presented in spite of the fact that the action (left and right mouse clicks) was correct in the disappointment acquisition experiment, that is, whether or not it was a disappointment trial. If “x” is presented as a correct action, the process proceeds to S1304. If “◯” is presented, the process proceeds to End.

  S1304 is a step of switching which stage of the average of the expected disagreement signals is added according to the number n of waveform acquisition repetitions in the disappointment situation counted from the start of the experiment. Steps S1305 to S1308 are steps of averaging the disappointment signals as step 1, step 2, step 3, and step 4, respectively. For example, since the addition is performed every four trials this time, four waveforms are assigned to each stage. In step S1309, the number of disappointing trials is incremented.

  By performing the steps from S1301 to S1309 with the data of four subjects, it is possible to add and average 16 disappointments with the same number of repetitions at each stage. This time, in order to investigate the time series change of the waveform while securing the number of additions for each stage, it was classified into 4 stages every 4 trials in consideration of the number of subjects, the number of repeated experiments, and the like.

  FIG. 6 shows a time-series change of the dissatisfied signal component obtained by averaging the respective disappointment stages. FIG. 6A shows an addition average waveform of the disappointment signal at each stage, where the horizontal axis is time and the unit is msec, the vertical axis is potential and the unit is μV. The amplitude and latency of the disappointment signal are different for each stage. However, when the stages are compared, the tendency of the waveform change including the amplitude and the latency can be seen. Since the influence of the noise component is considered to be reduced by the averaging of about 16 times in each stage, it is possible to discuss the tendency of the waveform change. In order to see the tendency in detail, FIGS. 6B and 6C were prepared. (B) is a diagram in which the time-series change of the amplitude of the disappointment signal is plotted at each stage, the vertical axis is the potential and the unit is μV. It can be seen that the amplitude of the disappointment signal at the initial stage is as large as about 20 μV, and the amplitude decreases as the number of repetitions of disappointment increases. (C) plots the latency of the disappointment signal at each stage, where the vertical axis is time and the unit is msec. It can be seen that the latency is shortened from nearly 600 msec to nearly 450 msec as the number of repetitions increases as compared to the stage 1 disappointment signal. As described above, it has been clarified that the amplitude and latency of the disappointing signal greatly change even when the signal is repeated about 20 times.

  Thus, based on the phenomenon that the inventors discovered through experiments, the amplitude of the disappointment signal decreases and the latency decreases as the number of disappointments increases, the rough usability evaluation shown in FIG. Details of each step of the processing flow will be described. The evaluation target device will be described using a DVD recorder as an example.

  First, FIG. 8 shows details of the processing in step S20 of FIG.

  First, a device operation instruction is issued from the usability evaluation apparatus. For example, in the case of a DVD recorder, instructions such as “please play”, “record” and “stop” are given (step S21).

  In response to this operation instruction, the user performs device operation (step S22). For example, a remote control or an operation panel of the device may be operated, or options displayed on the screen may be selected using a keyboard. Implementation of this device operation is a user request.

  As a result of the operation, the device is operated on the device side (step S23). At this time, regardless of whether or not the user has correctly operated, the device operates as if the user operated. For example, when the playback button is pressed in the scene of “Please record”, the playback operation is performed. That is, the device determines the response content in response to the user's request and presents the response content to the user. The operation of the device itself can be referred to as presentation of response contents, and the place where the operation is performed can be referred to as a presentation unit.

  The user observes how the device operates after operating his / her device. The brain wave of the user at this time is measured, and the potential change is sampled from the timing at which the operation from the device is presented in the specific event-related potential determination unit 18 (step S24). The user may be instructed to pay attention to the device operation, but even if there is no specific instruction, it can be considered that attention is paid to the reaction after the device operation as an operation of daily life.

Next, it is determined whether the user's operation is correct (step S25), and is stored in the storage unit 17 together with the electroencephalogram signal (step S26).
Since the data collection of one trial is completed by the processing so far, it is determined whether all scheduled operation instructions have been completed (step S27), and in the case of completion (YES in step S27) In step S28, if not completed (NO in step S27), the next function is selected (step S29), and the process returns to the first step S21.

  If YES in step S27, it means that data collection for a certain subject has been completed, so it is determined whether the experiments for all scheduled subjects have been completed (step S28). If YES in step S28), the data collection process (step S20) is complete, and if not completed (NO in step S28), the next subject is substituted and the process returns to the first step S21.

  By collecting data using such a loop, it is possible to acquire event-related potentials during operation for various functions for various conditions of the subject.

  After all the data is collected and step S20 is completed, next, the process for detecting the presence or absence of an unexpected signal that is a specific event-related potential is performed according to the flow shown in FIG. 9 (step S40).

  In FIG. 9, the data of the electroencephalogram (event related potential) is taken out from the storage unit 17 (step S41). Examples of the sampling frequency include 200 Hz, 500 Hz, and 1000 Hz. Next, characteristic time-domain sampling data that can be used for waveform identification is cut out from the reference time (starting point) of motion presentation (step S42). The waveform of the electroencephalogram near the reference time appears in response to auditory or visual stimuli and appears to have no relation to human expectation. Therefore, the portion is removed, for example, around 600 msec or 750 msec. Before and after, or cut from 200 msec to 800 msec, and used for determination.

  In step S43, noise is removed from the waveform formed by the extracted sampling data. For noise removal, in order to cut off high frequency components mixed in the signal, for example, through a 40 Hz low-pass filter, or to reduce the influence of electrooculosis caused by blinking, a waveform having an amplitude of 80 μV or more is removed from the identification target. Or

  Next, a standard waveform (for example, the graph 105 in FIG. 4; hereinafter referred to as a target template) that is created in advance and a normal waveform that is not disappointing is added (for example, the graph 105 in FIG. 4). The graph 104 in Fig. 4 (hereinafter referred to as a control template) and the correlation value between the cut out and denoised data waveform signals are calculated. That is, the correlation between the target template that symbolically represents the event-related potential waveform when the disappointment occurs and the collected data is calculated, and the normal event-related potential waveform that does not disappoint is represented symbolically The correlation between the control template and the collected data is calculated, and each correlation value is calculated. Based on the correlation value, it is calculated how much the waveform currently being processed in the collected data is correlated with the target template and how much is correlated with the control template (step S44).

  In step S45, a distance calculation is performed to determine which template is closest to each template and the currently processed waveform. Specifically, in a two-dimensional space that can be created from the correlation value between the target template of each waveform and the correlation value of the control template, calculate whether the waveform to be evaluated is closer to the control or target waveform set. To do. For the distance calculation, for example, Maharabinos distance is used. This Maharabinos distance indicates the distance from the center of gravity of the group taking into account the variance / covariance of the data, and is known to have higher discrimination ability than simple distance calculation.

  In step S46, it is determined whether the currently processed waveform is closer to the target or the control, based on a scale called Mahalanobis distance. If close to the target, the process proceeds to step S47, and if close to the control, the process proceeds to step S48.

  In step S47, since it is determined that the current waveform is closer to the target, the operator who has provided the current data identifies that the operator is in a state of disappointment. And it records in the storage part 17 that there is a disappointment.

  In step S48, since it is determined that the current waveform is closer to the control, the operator who provided the current data is not disappointing but identified that the response of the device as expected by the operator was obtained. To do. Then, it is recorded in the storage unit 17 as no disappointment.

  By following such a procedure, it is possible to discriminate to some extent even in an electroencephalogram, which is generally considered to be difficult to discriminate from one waveform to another due to large variations in waveforms.

  As described above, after all the event-related potential data stored in the storage unit 17 or a necessary amount of disappointment detection processing is performed, the function evaluation unit 16 determines the event-related potentials according to the flow shown in FIG. Adds waveform data. The added waveform obtained by this processing is used as data indicating whether or not the disappointing signal is decreasing.

  In FIG. 10, first, the data of the addition target waveform is extracted from the storage unit 17 (step S61). For example, if this is processing for a function of a single DVD player, only waveform data at the time of processing when the same function is performed is extracted. When it is desired to evaluate a specific person, only the data of the subject is extracted from the data of a plurality of persons.

  In step S62, a disappointment signal detection process is performed. If the detection process has already been performed in step S40, only the detection result may be used.

  In step S63, it is determined whether or not there is a disappointment signal. If a disappointment signal is detected (YES in step S63), the process proceeds to step S64. On the other hand, if a disappointment signal is not detected (NO in step S63), the process returns to step S61 and proceeds to extraction of the next waveform data.

  In step S64, the disappointing signal is added data at each stage of steps S65 to S68 depending on how many times the disappointing signal is detected from the start of data acquisition. Sort as. The distribution is realized by distributing each stage in a certain number, for example, two. That is, the first and second times can be added as a stage 1 waveform (step S65), and the third and fourth times can be added as a stage 2 waveform (step S66). In addition, when it can be assumed in advance how many disappointing signals are generated in the entire experiment, the number may be divided into four. For example, if a total of about 8 disappointing signals are likely to be generated, it may be decided to move to the next stage by adding two each.

  In step S69, it is determined whether or not all target waveform data to be subjected to addition processing has been processed. If all processing is completed (YES in step S69), the entire processing is terminated, and all processing is not completed. If so (NO in step S69), the process returns to step S61.

  In this way, it is possible to perform the process in the previous stage for determining the degree of decrease in the disappointment signal in the next step S80.

  Note that all target waveform data may be first extracted from the storage unit 17 to create a waveform data list, and thereafter, the processing from step 62 onward may be sequentially performed from the first acquired waveform data.

  Further, the data stored in the storage unit 17 includes the event-related potential waveform data itself based on the response of the device reaction unit 13 to the operation of the user 10, and an unexpected signal determination process (step S40) is basic data as biological signals such as presence / absence of a disappointing signal, amplitude and latency at the peak of the waveform, and what operation the basic data has performed for which operation instruction. , And is recorded in association with the action data.

  Further, after the processing in step S60 is performed, the storage unit 17 also stores data after each operation instruction and after addition processing for each operation. This is because the waveform when added at each stage, whether the added waveform contains a disappointing signal, the amplitude at the peak of the added waveform, and the latency as behavioral data as basic data as a biological signal Recorded in association with.

  This process is used when it is not possible to determine a decrease in waveform unless addition is performed, or when it is necessary to determine a decrease in waveform over a large number of times. In some cases, or when there is sufficient discrimination performance even with a non-addition waveform, this processing is not necessarily required.

  The user evaluation in step S80 is performed using the data stored in the storage unit 17 by the above processing. Specific usability evaluation processing is shown in FIG. This process corresponds to the specific content of the process in step S80, and is mainly processed by the function evaluation unit 16 which is a user evaluation unit.

  In usability evaluation, first, a function to be evaluated by the DVD player is selected (step S181), and the waveform data of the electroencephalogram measured when the user 10 operates the function and the detection result of the disappointment signal are extracted from the storage unit 17. Then go to the next process. The detection result of the waveform data and the disappointment signal is the same number of data as the number of times when the user 10 operates the function to be evaluated at present. The detection result of the disappointment signal includes the above-described information such as the presence / absence of the disappointment, peak amplitude, and latency.

  In step S182, it is determined whether the expected signal has been observed in the waveform data extracted in step S181 and the detection result of the expected signal. If a disappointment signal is not observed (NO in step S182), it can be understood that the function is already remembered when the same operation is performed several times. In step S185, it can be determined that it is simple. On the other hand, if a disappointment signal has been observed (YES in step S182), the process proceeds to the next step S183.

  In step S183, the degree of time-series change of the disappointment signal for a plurality of times is observed. This change is determined using the latency and amplitude data of a plurality of disappointment signals extracted from the storage unit 17. Judgment is based on the threshold processing of the amount of change, whether the latency and amplitude fall within a certain amount of change (within the fluctuation range of the measurement and experimental conditions), or whether there is a decreasing trend (measurement and experimental conditions Is the change more than fluctuation)? In addition to the threshold processing, when a typical decrease pattern is known in advance, it can be specified how much the pattern is decreased by comparison with the decrease pattern. The degree of this change is determined in step S184. If it is determined that the amplitude of the positive peak potential of the disappointing signal is decreasing (YES in step S184), the function is an initially unknown function, and thus a large disappointing signal is generated. After that, the magnitude of the disappointing signal decreased, so even if the same mistake was repeated, it was thought that it was gradually getting used and learning of the function operation was over, which is a new function for the subject. However, it can be determined that the function is easy to remember (step S186). For a function that was remembered once, a disappointing signal was observed only for the first time, and for a function that was memorized several times, this corresponds to a case where the disappointing signal was not detected in the first few times. . On the other hand, if the amplitude of the positive peak potential of the disappointment signal does not decrease (NO in step S184), the same operation error is caused in response to the same operation instruction, and a large disappointment signal continues to be generated. From this point, it is considered that there is a factor causing the same mistake over and over, and it is determined that the operation of the function is difficult to remember (step S187).

  After the determination is made in the above steps S185 to S187, the evaluation of the ease of remembering the function selected in step S181 is completed in any case, so the evaluation result in step S188. Record in the database. The database of evaluation results is in the function evaluation unit 16, and an example thereof is shown in FIG. In the table of FIG. 12, for each function, there is a disappointment signal (in the figure, “−” indicates that there was no disappointment signal), and if so, the amplitude and latency are recorded. Finally, the evaluation result is recorded. Details of the evaluation will be described later.

  By the processing so far, the evaluation of one function is completed, and in step S189, it is determined whether the evaluation of all functions is completed. If the evaluation of all functions is completed (YES in step S189), the process ends. If the evaluation of all functions is not completed (NO in step S189), the next function is selected. The process moves to step S181 for evaluation.

  Based on the above flow, an example of evaluating the degree of memorizing the operation of each function of the user's DVD player and determining the ease of remembering each function will be described with reference to FIG. This is also an example of a database of evaluation results in step S188. Here, there are a function A, a function B, and a function C, and the detection result of each disappointing signal is shown in the table. The inventors have discovered from experiments on the behavior when the disappointment signal continues multiple times, when the degree of disappointment decreases, the amplitude decreases and the latency increases. From this point of view, the function A corresponds to a case in which the disappointment was initially generated, but decreased in the meantime, and finally the disappointment signal itself was not detected. Therefore, in the flowchart of FIG. This can be considered as an example in which the usage of step S186 is easy to remember. It can be evaluated that the user has learned how to use it early.

  As for the next function B, since a disappointing signal is generated every time and no characteristic change is observed in the amplitude and the latency, a function that is not yet remembered and can be mistaken as in step S187 in FIG. It has been determined that it is difficult to remember. It can be evaluated that the user has not yet learned how to use it.

  For function C, the disappointing signal itself has not been detected from the beginning. The disappointment signal is a signal that occurs when the device operation imagined by the user deviates from the actual device operation. If this signal is not seen from the beginning, it can be used as expected from the beginning. The situation of step S185 in FIG. 11 is considered. It can be evaluated that the user is familiar with the state from the beginning.

  In this way, usability evaluation of a device such as a DVD player can be performed by using a change in the disappointment signal. In addition, by summing up by function, it is possible to determine which functions are difficult to use, more difficult to remember, and easy to make the same mistake.

  In addition, there are functions that the user has already remembered how to operate and those that have not yet been remembered, and it is considered that there are functions that are easy to remember and functions that are difficult to remember even for functions that are not remembered. This is considered to be caused by complicated factors such as the amount of knowledge of the user's device, experience so far, and operability imagined from the device interface.

  Because of these factors, when a user operates a device, a certain function can be operated without any problems for the first time, and a mistake is repeated even though a certain function has been operated several times. There is. It is necessary to determine the ease of remembering based on the degree of change of the event-related potential in the error repetition process, and to design an interface with few error repetitions.

  Especially in the elderly, as shown in Non-Patent Document 3, it is known that the same error is repeated in the same situation. By the evaluation method of the present invention, it is easy to remember in advance and it is possible to reduce error repetition. Realization of an interface is desired.

  In this way, by evaluating the decrease in the amplitude and latency of the disappointment signal and the degree of the decrease, it is possible to evaluate each function, not the entire device. Furthermore, it is possible to determine whether each function is easy to remember or difficult to remember, instead of determining whether each function can be used or not. This is an item that could not be measured in the past, and it is possible to find in advance an uncomfortable or unintuitive interface that may occur even if a user who cannot be distinguished by behavioral evaluation can operate the device. Although the evaluation can be considered as the degree of stress and relaxation as a whole by using physiological indices, according to the measurement using the event-related potential of the present invention, the number of trials of about several trials of each function Even when there are few, it can identify ease of remembering.

  In this embodiment, only the decrease in the amplitude and latency of the disappointment signal is used, but the method of the decrease is the same as the number of times this time, as in the present invention. In addition to what happens depending on whether or not it has happened, the inventors have found that the state of decrease differs depending on whether the disappointing situation has happened for the first time in a long time or continuously. When the law of decrease of the disappointing signal is clear, it becomes possible to make a more detailed disappointment determination and calculate the intensity of the disappointing signal. This corresponds to the refinement of step S40 in the flowchart of FIG.

(Embodiment 2)
In the second embodiment, an example of evaluating software that recommends information and services in an information device will be described.

  That is, in the first embodiment, the evaluation target is how easy the user can remember the function of the device, but it is not easy to remember the function but the user for the function provided by the device such as recommended software. It is assumed that the degree of interest of the person is evaluated. Since the disappointment signal is basically a signal calculated using the event-related potential, a plurality of evaluation items can be set depending on which event is related.

  The feature of this embodiment is that, when the amplitude or latency of the disappointment signal decreases, the user determines that he is bored or no longer interested in the recommended content of the software. In the information recommendation software, it is known that the user does not use the recommended software when the recommended content of the software is not useful to the user. In this case, a recommendation that is optimal for the user is required, as well as a content that is better than what the user expects, whether explicitly or implicitly. This is because even if the user already recommends content that is already known, it is not useful for the user and it is always necessary to introduce new content.

  For example, when using a car navigation system, etc., if the user requests that you introduce a nearby restaurant, no matter how much the user likes Chinese food, recommending the same Chinese restaurant every time will give new information to the user. This is a case where there is no fresh surprise. Also in this case, for example, when Thai food is recommended, it is considered that a disappointing signal can be detected in the sense that it is more than expected.

  Hereinafter, an embodiment of such a recommendation function evaluation will be described in detail with reference to FIGS.

  FIG. 13 shows the configuration of the usability evaluation apparatus according to the present embodiment, and the operation thereof will be described below with reference to FIG.

  In FIG. 13, 10 is a user, 21 is an operation input unit, 22 is a device reaction unit (presentation unit), 24 is a biological signal detection unit, 25 is a user evaluation unit, 26 is a storage unit, 27 is a parameter adjustment unit, and 28 is A specific event-related potential determination unit 29 is a response content determination unit. Reference numeral 23 denotes an information recommendation device that realizes an information recommendation function in which the operation input unit 21, the device reaction unit 22, and the response content determination unit 29 are integrated.

  When the user 10 wants to recommend some information to the information recommendation device 23 (for example, what is a delicious restaurant near here, what music is suitable for this season?) Etc.), a request is input to the operation input unit 21. This input can be, for example, voice, remote control, keyboard or mouse input.

  The information recommendation device 23 selects, in the response content determination unit 29, information that is considered to be most appropriate for the request from the user input through the operation input unit 21 (determines the response content), and the device reaction unit A response is made to the user 10 via 22 (presentation of response content). The device reaction unit 22 responds to the user 10 by, for example, voice or text or an image displayed on the display.

  When the information from the device reaction unit 22 is viewed, the biological reaction of the user 10, particularly the event-related potential measurable with an electroencephalograph is detected by the biological signal detection unit 24, and the event-related potential data stored in the storage unit 26 is detected. And transmission to the user evaluation unit 25. In addition, the event-related potential data is extracted from the storage unit 26, the presence / absence of the specific event-related potential is determined by the specific event-related potential determination unit 28, and the determination result is recorded in the storage unit 26.

  The user evaluation unit 25 determines an evaluation of how the current reaction of the device reaction unit 22 is with reference to a series of past reactions accumulated in the accumulation unit 26 so far, and as a result, a response content determination unit 29 is sent to a parameter adjustment unit 27 that adjusts parameters used for information recommendation. The response content determination unit 29 determines the response content to be presented to the user based on the user request and the parameters adjusted by the parameter adjustment unit 27.

  In this way, each time a user issues a request for information recommendation, the user evaluation unit 25 sets appropriate information recommendation parameters based on the reaction.

  Here, parameters in the information recommendation function will be described. For example, in order to realize an information recommendation function called a restaurant recommendation function, information such as the current user position, recommendable restaurant information, user preference, and past recommendation history is required. In order to recommend information, a distance scale is required to indicate how close a restaurant is to the user's preference at a distance from the current position of the user. By adjusting this distance measure, different restaurant information is provided to the user. Specifically, the distance scale is, for example, the restaurant that best suits the user's preference is 30 minutes by car from here, but the closest is a restaurant that is 5 minutes by car. The visit history does not exist. Here, adjusting the movement distance or the user's preference is an example of parameter adjustment. For example, if a parameter is adjusted so that the user's preference is emphasized, a restaurant that matches the user's preference is selected even if it is far away, and if a parameter is adjusted so that the movement distance is emphasized, the nearest restaurant is selected. It is said that it is done. Furthermore, there is a distinction according to the genre of food. There are also genres of menus with similarities in Japanese, Western, and Chinese genres, and how much importance is given to the distance between these menus is an example of parameter adjustment.

  By adjusting the parameters in this way, the user can always be provided with information by various recommendation methods.

  Next, a flowchart of the information recommendation evaluation process is shown using FIG. This process corresponds to the user evaluation process in step S80 in FIG. 7, and mainly corresponds to the internal process of the user evaluation unit 25 and the parameter adjustment unit 27 based on the evaluation.

  First, in response to a request through the operation input unit 21 of the user 10, optimal information is retrieved or generated from the situation where the user 10 is placed and presented to the user 10 through the device reaction unit 22 (step S281).

  In step S282, the event-related potential of the user's brain wave whose information is viewed by the biological signal detection unit 24 is acquired and measured, the specific event-related potential determination unit 28 determines whether a disappointment signal is included, and the disappointment signal Is determined not to be included (NO in step S282), information is provided as expected, and it is determined that the user 10 is not novel and tired (step S285). . As a result, the parameter adjustment unit 27 greatly changes the parameter used for recommendation (step S286), and at the next recommendation, information content greatly deviating from the similarity criterion currently used can be provided.

  If it is determined in step S282 that a disappointment signal is included (YES in step S282), it is confirmed that the user has caused an unexpected situation.

  In the next step S283, it is determined what tendency the disappointment signal has compared with the disappointment signals so far. If it is determined that the intensity, size, etc. tend to decrease as compared to the disappointing signal so far (YES in step S284), the user may be bored using the same parameters. It is determined that the degree of interest has decreased (step S287), and the parameter adjustment unit 27 slightly changes the parameters of the recommended engine (step S288). For example, this corresponds to providing information content slightly deviating from the similarity standard.

  Such parameter adjustment can be made by simple judgment based on the presence / absence of a disappointing signal. However, according to the present invention, any parameter can be selected depending on the strength of the decreasing tendency such as the intensity or magnitude of the disappointing signal. It is characterized in that it can be adjusted whether the parameter should be changed. For example, this is effective when disappointment is observed but the amplitude gradually decreases. If only the presence / absence of disappointment is used, disappointment is still observed, and no response can be made at this point. In particular, in terms of interest in the recommended software, it is considered difficult to regain interest in the recommended software when the interest is completely lost.

  If a disappointment signal is observed at the same level as before in step S284 (NO in step S284), the user has a disappointment signal that leads to a fresh surprise for the presented information, It can be determined that the user is not bored (step S289), and in the next step S290, it is not necessary to change the parameter in the parameter adjustment unit 27, and the process ends without changing the parameter of the recommendation engine.

  In this embodiment, it can be said that how much the user 10 has the expectation or interest in the recommendation information is the state of the user 10 that the user evaluation unit 25 evaluates.

  The flowchart of FIG. 14 shows the process for the parameter correction method for a single information presentation, and this process is repeatedly performed during actual operation. At this time, as in the storage unit 17 in the first embodiment, the storage unit 26 presents information regarding what kind of request together with biological data such as an event-related potential waveform, presence / absence of an unexpected signal, amplitude, and latency. In addition, data is recorded in association with the situation data indicating how much disappointment signals were observed.

  By performing evaluation based on such a flow, it is possible to always make recommendations based on appropriate parameters in the information recommendation device.

  In addition, the recommendation algorithm itself can be evaluated. An example is shown in FIG. This is a case where the recommendation software A and the recommendation software B are evaluated. In this case, since the disappointing signal related to the user's surprise is maintained in the recommended software B, the information recommending software B is less likely to get bored by the user, so it can be determined that the performance is high. Here, the recommendation software A and the recommendation software B are regarded as specific functions in the information recommendation device.

  In this way, unlike the first embodiment, in this embodiment, it is determined that the degree of interest of the user is sustained by considering that the intensity or size of the disappointing signal is not good. You can evaluate the software that makes recommendations. In addition, parameter adjustment can be performed by associating whether or not the parameter for information presentation of the recommended software is appropriate. As a result, it is possible to always realize a fresh and surprising recommendation by appropriately adjusting the recommendation software that has been bored with simple recommendation contents.

(Embodiment 3)
In the present embodiment, an embodiment will be described in which the application is not the device evaluation but the evaluation of each individual user (evaluation of the user state). Since the present invention uses a brain wave when the user operates the device, it can be used comprehensively for evaluating the device, and conversely, it can be applied to evaluating how much and what function the user can use.

  In general, the ease of use of a device has a part that does not depend on the user such as button arrangement, color, and size, and a part that greatly depends on the user's previous experience, such as menu configuration and operation method. In particular, since personal computers, DVD recorders, mobile phones, car navigation systems, etc. have functions that can operate various information, there are various forms of information operation methods. The operation system is very different.

  For this reason, as the user becomes more familiar with the method of using the device that has been used, the difficulty in using the device of another type or manufacturer may increase. For example, taking a mobile phone as an example, Mr. A who used A1's mobile phone A1 and Mr. B who used B's mobile phone B1 switched to A's mobile phone A2 at the same time. For example, for Mr. A, the mobile phone A2, which has an operation system similar to that of the mobile phone A1, seems to be easy to use. For Mr. B, the mobile phone A2 makes a phone call, takes a picture, mail Despite the same functions such as sending messages, there are slight differences such as mode switching, character input method, how to use the phone book, etc., depending on the operation method of the mobile phone B1 used so far, a new mobile phone A2 It is assumed that the same mistakes are repeated to operate.

  In this way, even in the evaluation of the same device, the evaluation varies greatly for each individual. Therefore, this embodiment evaluates the proficiency level of the function for each individual as the state of the user.

  FIG. 16 shows the configuration of the usability evaluation apparatus in the present embodiment.

  In FIG. 16, 10 is a user, 31 is a problem display unit, 32 is an operation input unit, 33 is a correct / incorrect display unit (presentation unit), 35 is a biological signal detection unit, 36 is a user evaluation unit, 37 is a storage unit, and 38 is The specific event-related potential determination unit 39 is a response content determination unit. The understanding level determination device 34 has a function of integrating the problem display unit 31, the operation input unit 32, the correct / incorrect display unit 33, and the response content determination unit 39 to present and evaluate an evaluation target question (question).

  The understanding level determination device 34 sequentially presents the problem to the user 10 via the problem display unit 31, and the user 10 inputs the answer via the operation input unit 32 (user request input). In the understanding level determination device 34, the response content determination unit 39 determines whether the answer from the user 10 is correct, and displays whether the answer is correct or incorrect to the user 10 via the correct / incorrect display unit 33. The loop so far has the same configuration as the learning support device, which automatically issues a question and answers the correctness.

  In the present embodiment, not only how much the user 10 understands at this time but also the determination of the proficiency level of the user 10 is performed by the biological signal detection unit 35 so that the user 10 A biological signal when the reaction is received, particularly an event-related potential measurable by an electroencephalograph, is detected, and event-related potential data is stored in the storage unit 37. The specific event-related potential determination unit 38 reads event-related potential data from the storage unit 37, determines the presence / absence of a specific event-related potential, and records the determination result in the storage unit 37.

  The user evaluation unit 36 performs an evaluation based on the detected biological signal and past data stored in the storage unit 37. The user is familiar with which problem, and for which problem It is a system that can determine whether you are not yet proficient.

  In the configuration of the present invention, all the problems (questions) are presented and the answers are received, and the offline operation that is evaluated after accumulating all the event-related potential data at that time is described. Online evaluation can be performed by evaluating the user evaluation unit 36 after answering one question. In this case, if the user's evaluation unit 36 determines that the proficiency level of a certain problem is insufficient, it is possible to take online countermeasures such as improving the question frequency of the problem.

  The problem here refers to, for example, an operation instruction in device evaluation as a problem.

  FIG. 17 is a flowchart of the proficiency level evaluation process according to this embodiment. This process corresponds to the user evaluation process in step S80 in FIG. 7, and is mainly processed inside the user evaluation unit 36.

  First, a function to be evaluated in the device is selected (step S381), and a waveform of an electroencephalogram when the device is operated with respect to the function is extracted from the storage unit 37. The result of the extraction is the same number of data as the number of times when the user operates the function to be evaluated at present when the averaging is not performed, and the number of stages set at step S60 when the averaging is performed. The same number of data.

  In step S382, the user evaluation unit 36 determines whether a disappointment signal is observed in the detection result of the event-related potential data extracted in step S381. If a disappointment signal has never been observed (NO in step S382), even if the same function is evaluated several times, there is no disappointment signal, and it can be understood that the function is already remembered. It can be determined that the function is simple in step S385. On the other hand, when a disappointment signal is observed (YES in step S382), the process proceeds to the next step S383.

  In step S383, the degree of time-series change of the disappointment signal for a plurality of times is observed. Possible changes include a case where no characteristic change is observed in the amplitude and latency, and a case where it is determined that the amplitude and latency are gradually decreasing. The degree of this change is determined in step S384. If it is determined that the amplitude of the positive peak potential is decreasing (YES in step S384), it was a problem that was not known at the beginning, and thus a large disappointing signal was generated. Since it has decreased, even if the same mistake is repeated, it can be determined that learning is gradually progressing and learning is progressing (step S386). For a function that was remembered once, a disappointment signal was observed only for the first time, and for a function that was remembered a few times, it corresponds to the case where the disappointment signal was not detected in the first few times To do. On the other hand, if the amplitude does not decrease (NO in step S384), it is considered that the same problem is caused by a similar operation error, and a large disappointment signal continues to be generated, so that it is not yet accomplished. It is determined that it has not been used yet (S387).

  After the determination is made as in steps S385 to S387 above, in any case, the evaluation of the ease of remembering the function selected in step S381 (the evaluation of the proficiency level of the specific user for the selected function) ) Is completed, it is recorded in the evaluation result database in step S388. Similar to step S188 of the first embodiment, this evaluation result database is installed inside the user evaluation unit 36 and has a format similar to that shown in FIG.

  With the processing so far, the evaluation of the user's proficiency level for a certain function is completed, and it is determined in step S389 whether the evaluation for all functions is completed. If the evaluation of all functions is completed (YES in step S389), the process ends. If the evaluation of all functions is not completed (NO in step S389), the next function is selected. The process moves to step S381 for evaluation.

  In this way, the degree of understanding and proficiency of each function of the user can be determined from the degree of decrease in the amplitude of the disappointment signal. For example, this also clarifies the difference in the proficiency level for each function of the user group of the mobile phone A and the user group of the mobile phone B. Based on the determination result, menus and help that are easy to remember for both groups are designed. You can also prepare a mechanism such as switching menus and help by entering the model of the mobile phone used in advance.

  In the present invention, the method of summing up after collecting data has been described. However, as in the second embodiment, the event-related potential can be acquired online and changed immediately. .

(Other embodiments)
The embodiments described so far are examples of the present invention, and the present invention is not limited to these examples. For example, as shown in FIG. 18, a specific event-related potential determination unit 18 may be provided between the biological signal detection unit 15 and the storage unit 17 in the apparatus configuration according to the first embodiment. In this case, the event-related potential data measured by the biological signal detection unit 15 is determined by the specific event-related potential determination unit 18 for the presence or absence of the specific event-related potential, and is sent to the storage unit 17 as it is. Accumulated. A similar device configuration may be applied to the second and third embodiments.

  In addition, the user state such as the degree of user's memorizing operation, the degree of interest, and the proficiency level of the operation can be determined by changing, for example, decreasing the amplitude of the waveform of the disappointment signal that is a specific event related potential. Although the determination method for performing the determination in the evaluation unit has been described so far, the determination may be performed using the peak latency of the waveform of the disappointing signal or the shape of the waveform as a determination criterion in addition to the amplitude of the waveform.

  The usability evaluation apparatus according to the present invention has a user behavior evaluation unit that evaluates user behavior based on a time-series change in event-related potentials that can be measured with an electroencephalograph. This is useful for evaluating time-series changes such as evaluation.

It is a block diagram of the usability evaluation apparatus which concerns on Embodiment 1. FIG. It is a conceptual diagram of apparatus operation. It is a figure which shows the procedure of a preliminary experiment. It is a figure which shows the acquisition experiment result of a disappointment signal. It is a figure which shows the analysis method of a disappointment signal. It is a figure which shows the time-sequential change of a disappointment signal. It is a figure which shows the flowchart of the whole process. It is a figure which shows the flowchart of experiment data collection of Embodiment 1. FIG. It is a figure which shows the flowchart of a disappointment detection process of Embodiment 1. FIG. It is a figure which shows the flowchart of the waveform addition process of Embodiment 1. FIG. It is a figure which shows the flowchart of the function evaluation process of Embodiment 1. FIG. It is a figure which shows the example of the potential change assumed. It is a block diagram of the usability evaluation apparatus which concerns on Embodiment 2. It is a figure which shows the flowchart of a user evaluation part. It is a figure which shows the example of the potential change assumed. It is a block diagram of the usability evaluation apparatus which concerns on Embodiment 3. It is a figure which shows the flowchart of the user evaluation part which concerns on Embodiment 3. FIG. It is a different block diagram of the usability evaluation apparatus which concerns on Embodiment 1. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 User 11 Operation instruction | indication part 12 Operation input part 13 Apparatus reaction part 14 Evaluation object apparatus 15 Biosignal detection part 16 Function evaluation part 17 Accumulation part 18 Specific event related electric potential determination part 19 Response content determination part 21 Operation input part 22 Equipment reaction part DESCRIPTION OF SYMBOLS 23 Information recommendation function 24 Biosignal detection part 25 User evaluation part 26 Accumulation part 27 Parameter adjustment part 28 Specific event related electric potential determination part 29 Response content determination part 31 Problem display part 32 Operation input part 33 Correct / incorrect display part 34 Evaluation object problem 35 Living body Signal detection unit 36 User evaluation unit 37 Storage unit 38 Specific event related potential determination unit 39 Response content determination unit

Claims (14)

An operation input unit that receives a user request;
A response content determination unit for determining a response content with respect to the request received by the operation input unit;
A presentation unit for presenting the response content determined by the response content determination unit to the user;
A biological signal detector for measuring an event-related potential of an electroencephalogram as the user's biological signal;
An accumulator for accumulating the measured event-related potential data;
Starting from the timing at which the presentation unit presents the response content, the presence or absence of a specific event-related potential after a predetermined time from the starting point is determined from the event-related potentials measured by the biological signal detection unit. A specific event-related potential determination unit that accumulates the determination result in the storage unit;
In the data of a plurality of the event-related potentials stored in the storage unit and having the specific event-related potential, the state of the user is evaluated from the degree of time-series change of the plurality of specific event-related potentials. A usability evaluation device comprising: a user evaluation unit that performs:   The usability evaluation apparatus according to claim 1, wherein the user evaluation unit evaluates a user state based on a time-series change in a peak amplitude of a waveform of the specific event-related potential.   The usability evaluation apparatus according to claim 1, wherein the user evaluation unit evaluates a user state based on a time-series change in a peak latency of a waveform of the specific event-related potential.   The usability evaluation apparatus according to claim 1, wherein the user evaluation unit evaluates a user state based on a time-series change in a waveform shape of the specific event-related potential.   The user evaluation unit evaluates the degree to which the user learns the operation of the function when the user operates the specific function of the predetermined device a plurality of times as a user state, and learns the operation of the function. Ease of use is determined, The usability evaluation device according to any one of claims 1 to 4.   The user evaluation unit compares each degree of time-series change of the specific event-related potential in the operation of two functions, and the function operation is performed when the amplitude of the waveform of the specific event-related potential decreases earlier. The usability evaluation apparatus according to claim 5, further determining that it is easy to remember.   The usability evaluation apparatus according to any one of claims 1 to 4, wherein the user evaluation unit evaluates the degree of interest of the user with respect to a specific function provided by a predetermined device as a user state. .   The user evaluation unit compares the degree of time-series change of the specific event-related potential when the two specific functions are provided, and the degree of decrease in the amplitude of the waveform of the specific event-related potential The usability evaluation apparatus according to claim 7, wherein it is determined that the interest is maintained when the number is smaller.   The usability evaluation apparatus according to any one of claims 1 to 4, wherein the user evaluation unit evaluates a user's proficiency with respect to an operation of a specific function in a predetermined device as a user state.   The usability evaluation apparatus according to claim 9, wherein the user evaluation unit determines that the user has mastered an operation of a specific function when an amplitude of a waveform of the specific event-related potential decreases.   The usability evaluation apparatus according to any one of claims 1 to 10, wherein the predetermined time is 450 to 800 msec.   The usability evaluation apparatus according to any one of claims 1 to 10, wherein the predetermined time is approximately 600 msec or approximately 750 msec. Receiving a user request;
Determining a response content for the request;
Presenting the response content to the user;
Measuring an event-related potential of an electroencephalogram as the user's biological signal;
Storing the measured event-related potential data;
Starting from the timing of presenting the response content as a starting point, determining whether or not there is a specific event related potential after a predetermined time from the starting point, and accumulating the determination result ,
Evaluating the state of the user from the degree of time-series changes of the plurality of specific event-related potentials in the accumulated data of the plurality of event-related potentials having the specific event-related potential. Evaluation methods. Receiving a user request;
Determining a response content for the request;
Presenting the response content to the user;
Measuring an event-related potential of an electroencephalogram as the user's biological signal;
Storing the measured event-related potential data;
Starting from the timing of presenting the response content as a starting point, determining whether or not there is a specific event related potential after a predetermined time from the starting point, and accumulating the determination result ,
Evaluating the state of the user from the degree of time-series changes of the plurality of specific event-related potentials in the accumulated data of the plurality of event-related potentials having the specific event-related potential; A program to be executed.

Images