JP2005245764A - Inference information preparation apparatus, inference information management system, and inference information preparation program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve reliability of inference information prepared based on information measured from a sensor and to expand its utilization range by clarifying the source of an inference engine for inferring an index value different from a measured value. <P>SOLUTION: In an inference information preparation device 1, measured values from the respective sensors are obtained connected with a temperature sensor 182 for measuring the body temperature of a user, a sweating sensor 183 for measuring sweating, and a heartbeat rate sensor 184 for measuring a heartbeat rate through an input detection part 180. By the inference engine having an inference engine ID being a characteristic ID number, inference data being an inference result based on the measured values from the respective sensors is prepared. The inference engine ID is imparted to the prepared inference data to prepare the inference information about the consciousness, emotion, etc. of the user. Processing according to characteristic is carried out to turn the inference information to be data convenient or profitable for using after that. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to an inference information creation apparatus, an inference information management system, and an inference information creation program for creating inference data that is an index value different from a measurement value by an inference engine based on information measured from a sensor.

  2. Description of the Related Art Conventionally, various devices have been proposed that provide a sensor for measuring a user's physiological information, biological information, and the like, and infer the user's consciousness and emotion based on various information measured by the sensor.

In addition to sensors that measure physiological information such as heart rate and skin impedance, a CCD camera for detecting the user's posture and motion state and a microphone for detecting the voice emitted by the user are provided. There is known a consciousness level detection device that can detect more accurately whether or not a person's consciousness level is in a specific state. Furthermore, a game control device is known in which a user can intentionally input a psychological state intentionally (see, for example, Patent Document 1).
Japanese Patent Laid-Open No. 10-57355

  By the way, when inferring based on various information measured from a sensor, it is known that the inference contents and inference results output differ depending on factors such as the inference method and setting conditions. The contents of these inference methods and setting conditions are usually different depending on the inference means (inference engine) included in each inference device.

  However, when the reasoning information created by each reasoning device is used later, if the source of the reasoning engine that created the reasoning information is unknown, it is difficult to maintain data consistency in the device at the point of use. In addition, the user may determine that the inference information is not reliable. These circumstances have been a major impediment to the use of inference information.

  The present invention has been made in order to solve the above-mentioned problems. The origin of an inference engine that infers an index value different from a measured value is clarified, and inference information created based on information measured from a sensor is obtained. An object of the present invention is to provide an inference information creation device, an inference information management system, and an inference information creation program that can increase the reliability and expand the range of use.

  In order to achieve the above object, an inference information creation device according to claim 1 includes a measurement value acquisition unit that acquires measurement values from at least one sensor, and the measurement acquired by the measurement value acquisition unit. An inference data creating means for creating inference data that is an index value different from the measured value based on a value; identification information adding means for adding identification information unique to the inference data creating means to the inference data; Inference information output means for outputting inference information including the inference data to which the identification information is added.

  In addition to the configuration of the invention according to claim 1, the inference information creation device of the invention according to claim 2 includes the identification information of the inference data creation means, characteristic information indicating characteristics of the inference data creation means, And a characteristic information acquisition means for acquiring the characteristic information corresponding to the identification information included in the inference information output by the inference information output means from the characteristic information table. And.

  According to a third aspect of the present invention, there is provided an inference information creating apparatus according to the second aspect of the present invention, wherein the characteristic information includes at least one of reliability, latest update date, and inference type of the inference data creating means. It is characterized by including one.

  In addition, the inference information creation device of the invention according to claim 4 includes at least one processing procedure executed for the inference information in addition to the configuration of the invention of claim 2 or 3, and the characteristic information acquisition means A processing procedure selection means for selecting one processing procedure from a plurality of the processing procedures based on the characteristic information acquired by the processing procedure, and the inference information based on the processing procedure selected by the processing procedure selection means. Inference information processing means for processing the inference information output by the output means.

  An inference information management system of an invention according to claim 5 includes an inference information creation device that creates inference information of a user based on measurement values acquired from at least one sensor, and the inference information creation device. An inference information management system that manages the inference information created by the inference information management system connected via a network, wherein the inference information creation device acquires the measurement value from the sensor An acquisition unit, an inference data generation unit that generates inference data that is an index value different from the measurement value based on the measurement value acquired by the measurement value acquisition unit, and an identification unique to the inference data generation unit Identification information adding means for adding information to the inference data, and inference information output means for outputting the inference information including the inference data to which the identification information is added The inference information management apparatus stores inference information acquisition means for acquiring the inference information output from the inference information creation apparatus via a network, and the inference information acquired by the inference information acquisition means The inference information storage means, the identification information of the inference data creation means, and the characteristic information table storing the characteristic information indicating the characteristics of the inference data creation means in association with each other, and output by the inference information output means Characteristic information acquisition means for acquiring the characteristic information corresponding to the identification information included in the inferred information from the characteristic information table.

  In addition to the configuration of the invention according to claim 5, the inference information management system of the invention according to claim 6 includes at least one of the reliability of the inference data creation means, the latest update date, and the inference type. It is characterized by including one.

  In addition to the configuration of the invention according to claim 5 or 6, the inference information management system of the invention according to claim 7 includes at least one processing procedure executed for the inference information. A processing procedure selection unit that selects one processing procedure from a plurality of processing procedures based on the characteristic information acquired by the characteristic information acquisition unit; and the processing procedure selected by the processing procedure selection unit And inference information processing means for processing the inference information output by the inference information output means.

  An inference information management system according to an eighth aspect of the present invention is the inference information management system according to any one of the fifth to seventh aspects, wherein the inference information output means communicates with the inference information management device wirelessly or by wire. 1st communication interface means for performing data transmission / reception, The said inference information acquisition means is provided with the 2nd communication interface means for performing data transmission / reception with the said inference information creation apparatus by radio | wireless or a wire communication. Yes.

  According to a ninth aspect of the invention, there is provided an inference information creation program that uses a measurement value acquisition unit that acquires a measurement value from at least one sensor and a measurement value acquired by the measurement value acquisition unit. Inference data creation means for creating inference data that is an index value different from the measured value, identification information addition means for adding identification information unique to the inference data creation means to the inference data, and the identification information is added And functioning as inference information output means for outputting inference information including the inference data.

  In the inference information creation device of the invention according to claim 1, inference data that is an index value different from the measurement value is created based on the measurement value acquired from each sensor, and identification information unique to the inference data creation means Inference information including inference data to which is added is output. Therefore, the origin of the inference data creation means can be clarified, and the reliability of the inference information created based on the information measured from the sensor can be improved.

  In addition, in the inference information creation device of the invention according to claim 2, in addition to the effect of the invention according to claim 1, the identification information of the inference data creation means and the characteristic information indicating the characteristics of the inference data creation means respectively correspond to each other A characteristic information table is provided for storing the characteristic information, and characteristic information of the inference data creating means that has created the inference information from the characteristic information table is acquired. Therefore, the origin and characteristics of the inference data creation means can be grasped.

  In addition, in the inference information creation device of the invention according to claim 3, in addition to the effect of the invention according to claim 2, the characteristic information includes the reliability of the inference data creation means, the latest update date, and the inference type. The source and characteristics of the data creation means can be grasped.

  In addition, in the inference information creation device of the invention according to claim 4, in addition to the effect of the invention according to claim 2 or 3, one of a plurality of processing procedures is selected based on the characteristic information, and processing is performed according to the processing procedure. Is executed. Therefore, processing according to the characteristics of the inference information is executed, and the use range of the inference information can be expanded.

  Further, in the inference information management system of the invention according to claim 5, inference information is collected in the inference information management device from the inference information creation device that creates inference information about the user, and the inference information management creation device generates the inference information. Characteristic information is acquired based on the included identification information. Therefore, the origin of the inference data creation means can be clarified, and the reliability of the inference information created based on the information measured from the sensor can be improved.

  Further, in the inference information management system of the invention according to claim 6, in addition to the effect of the invention according to claim 5, the characteristic information includes the reliability of the inference data creation means, the latest update date, and the inference type. The source and characteristics of the data creation means can be grasped.

  In addition, in the inference information management system of the invention according to claim 7, in addition to the effect of the invention according to claim 5 or 6, one of the plurality of processing procedures is selected based on the characteristic information, and processing is performed according to the processing procedure. Is executed. Therefore, processing according to the characteristics of the inference information is executed, and the use range of the inference information can be expanded.

  In addition, in the inference information management system according to the eighth aspect of the invention, in addition to the effects of the invention according to any one of the fifth to seventh aspects, the inference information creation device and the inference information management device each have an interface for transmitting and receiving data. Since the means is provided, it is possible to connect the inference information creation device and the inference information management device provided separately from each other via a network.

  Further, in the inference information creation program of the invention according to claim 9, inference data that is an index value different from the measurement value is created based on the measurement value acquired from each sensor, and is unique to the inference data creation means. Inference information including the inference data to which the identification information is added is output. Therefore, the origin of the inference data creation means can be clarified, and the reliability of the inference information created based on the information measured from the sensor can be improved.

  Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. The inference information creation device according to the first embodiment is a small portable terminal device carried by a user.

  In the inference information creation apparatus according to the present embodiment, the inference information is created by the inference engine based on the information measured from the sensor, and the inference information is created, and further processing according to the characteristic information Is executed. Below, each measured value of body temperature, perspiration, and heart rate is illustrated as information measured from a sensor. In addition, inference information here is information regarding a user's consciousness and emotion. In the present embodiment, it is assumed that the inference information relates to the “impression” of the user, and a case will be described in which inference information corresponding to the level of the “impression” is created.

  First, with reference to FIG. 1 thru | or FIG. 3, the structure of the inference information creation apparatus 1 which concerns on 1st Embodiment is demonstrated. FIG. 1 is a block diagram showing a configuration of the inference information creation apparatus 1. FIG. 2 is a conceptual diagram showing the configuration of the storage area of the RAM 130 of the inference information creation apparatus 1. FIG. 3 is a conceptual diagram showing the configuration of the storage area of the HDD 140 of the inference information creation apparatus 1.

  As shown in FIG. 1, the inference information creation apparatus 1 is provided with a CPU 110 that controls the inference information creation apparatus 1. The CPU 110 includes a ROM 120 that stores a program such as BIOS executed by the CPU 110, a RAM 130 that temporarily stores data, and a hard disk drive (hereinafter referred to as “HDD”) that is a data storage device. 140) is connected. In addition, a timing device 190 for counting the current date and time and the time interval is connected to the CPU 110 via the bus 115. The timing device 190 is an IC chip having a clock function. Further, the timing device 190 may be configured to acquire the date and time via the Internet or wirelessly.

  In addition, an input detection unit 180 that detects input from various devices is connected to the CPU 110 via the bus 115, and the input detection unit 180 allows a user to operate the inference information creation apparatus 1. An input panel 181 having buttons and switches, a body temperature sensor 182 for measuring the user's body temperature, a sweat sensor 183 for measuring the user's sweat state, and a heart rate for measuring the user's heart rate A sensor 184 is connected. The body temperature sensor 182, the sweat sensor 183, and the heart rate sensor 184 are not limited in position and measurement method as long as the body temperature, sweat, and heart rate can be measured effectively. A sensor reading unit is provided. The body temperature sensor 182 measures 0 to 50 ° C. as a measured value, the sweat sensor 183 measures 0 to 100% RH, and the heart rate sensor 184 measures 0 to 200 beats as a measured value. And The information (measurement value) measured from the body temperature sensor 182, the sweat sensor 183, and the heart rate sensor 184 corresponds to the “measurement value” of the present invention.

  Each sensor is controlled to automatically perform periodic measurement when the inference information creating apparatus 1 is powered on and activated. The measurement value of each sensor is stored in a predetermined storage area inside each sensor, and the inference information creation apparatus 1 acquires the latest measurement value from the predetermined storage area via the input detection unit 180. In addition, a measurement value storage area (not shown) for each sensor is provided in the RAM 130 or the HDD 140 of the inference information creation apparatus 1, and the measurement value of each sensor is stored in the measurement value storage area via the input detection unit 180, The inference information creation apparatus 1 may refer to this measurement value storage area and acquire the latest measurement value.

  Although not shown, the input panel 181 includes a power reset switch for turning the power of the inference information creation apparatus 1 on and off, and restarting the inference information, and when the inference information creation apparatus 1 includes a plurality of inference engines. An inference engine selection switch for the user to select an arbitrary inference engine and a characteristic-specific process instruction switch for executing a characteristic-specific process are provided.

  An inference engine has a function of inferring user's consciousness and emotion based on information measured from each sensor. In the inference engine, an inference method and setting conditions are defined. The inference engine is a program for inferring the user's consciousness and the like based on the measurement value from each sensor according to the definition contents, and is executed by the CPU 110. In this embodiment, as will be described later, a plurality of inference engines are stored in the HDD 140, and the user can select any inference engine.

  With this configuration, in the inference information creation device 1 according to the first exemplary embodiment, based on the sensor information from the body temperature sensor 182, the sweat sensor 183, and the heart rate sensor 184, the user's consciousness is inferred by the inference engine. Thus, inference information is created, and processing corresponding to the characteristic information is executed. Note that an inference information creation program is executed as one of the modules executed by the inference information creation apparatus 1 of the present embodiment. It is assumed that the inference information creation program is stored in advance in the program storage area 142 (see FIG. 3) on the HDD 140. Further, when installed via a CD-ROM drive, floppy (registered trademark) disk drive, and various interfaces (not shown), an external storage medium such as a CD-ROM or an external storage device via a network can be used on the HDD 140. An inference information creation program may be set up in the program storage area 142 and the information storage area 143 (see FIG. 3).

  As shown in FIG. 2, the RAM 130 of the inference information creation apparatus 1 stores a work area 131 that stores temporary data during execution of the program and an input information storage area 132 that temporarily stores various types of input information. And an output information storage area 133 for temporarily storing various information to be output. Further, the RAM 130 is provided with various storage areas not shown.

  As shown in FIG. 3, the HDD 140 of the inference information creation apparatus 1 has an operating system (OS) storage area 141 that stores various programs executed by the CPU 110 to control the operation of the inference information creation apparatus 1, and an inference A program storage area 142 that stores various programs executed by the information creation device 1 and an inference information creation program, an information storage area 143 that stores information such as settings, initial values, and data necessary for program execution, and creation There is provided an inference information storage area 144 for storing the inferred information, and a characteristic information storage area 145 for storing the inference information processed for each characteristic. The information storage area 143 stores an inference definition table and a characteristic information table which will be described later.

  In the program storage area 142, a plurality of inference engines are stored as a part of the inference information creation program. Each inference engine stores in advance an inference engine ID that is unique identification information as a constant, and each inference engine can be uniquely specified by this inference engine ID. The inference engine ID is basically ID information that cannot be rewritten.

  Hereinafter, the flow of processing in the inference information creation device 1 of the present invention will be described with reference to the drawings. In the inference information creation device 1, an “inference information creation process” for creating inference information based on measurement values from each sensor and an “inference information characteristic-specific process” for executing a process according to the characteristics are the main processes. Run as.

  First, the “inference information creation process” will be described with reference to FIGS. FIG. 4 is a main flowchart of the inference information creation process. FIG. 5 is a flowchart showing details of the inference engine initialization process (S1). FIG. 6 is a flowchart showing details of the inference data creation process (S3). FIG. 7 shows the data structure of the inference definition table 13. FIG. 8 is a flowchart showing details of the inference information output process (S11). FIG. 9 is a diagram illustrating a data configuration of the inference information 10.

  The main flowchart (FIG. 4) of the inference information creation process is the case where the user operates the power reset switch (not shown) of the input panel 181 to turn on the inference information creation device 1, or inference information creation. When the device 1 is reset, the processing is started.

  As shown in FIG. 4, first, an inference engine initialization process (S1) is executed. The inference engine initialization process is a process for initializing a reference value for each sensor that is referred to in the later-described inference data creation process.

  As shown in FIG. 5, in the inference engine initialization process (S1), “0” is assigned to the variable ST, variable SH, and variable SM (S101), and “3” is assigned to the variable T (S102). . And a measured value is acquired from each sensor (S103). That is, the body temperature sensor 182, the sweat sensor 183, and the heart rate sensor 184 measure the body temperature, sweat, and heart rate of the user, and the measurement values are acquired from these sensors.

  The measured values of body temperature, sweating, and heart rate acquired from each sensor are added to the variable ST, variable SH, and variable SM, respectively (S104). If the measurement value is acquired for the first time in S103, since “0” is substituted for the variable ST, variable SH, and variable SM in S102, each of the body temperature, sweating, and heart rate acquired in S103 is obtained. The measured value is directly substituted into the variable ST, variable SH, and variable SM.

  Then, the variable T is decremented by “1” (S105), and if T is not “0” (S106: NO), the process returns to S103, and the measurement values obtained by the sensors are acquired again. In this way, until the number of times set to the variable T in S102 (here, 3 times) is reached, the measurement value acquisition in S103 is executed, and the processing in S103 to S106 is repeated. As a result, if T is “0” (S106: YES), the variable ST, variable SH, and variable SM store the total value of the measurement values for the number of times set in the variable T.

  The variable ST, variable SH, and variable SM are respectively divided by “3”, which is the value set in the variable T, and assigned to the variable CT, variable CH, and variable CM, respectively (S107). That is, by dividing the total value of the measured values for each sensor by the number of times of measurement, an average value (= normal reference value) of one measured value of each sensor is obtained.

  The variable CT is a reference value for the body temperature sensor 182, the variable CH is a reference value for the sweat sensor 183, and the variable CM is a reference value for the heart rate sensor 184. The reference values of these sensors are stored in a reference value area (not shown) provided in the RAM 130.

  Returning to FIG. 4, the selection of the inference engine is executed (S2). The inference information creation apparatus 1 can select an arbitrary one from a plurality of inference engines, and the processing contents for creating inference data differ depending on each inference engine. Each inference engine is stored in the program storage area 142, and an inference engine for executing an inference data creation process described later is determined in S2.

  The selection of the inference engine (S2) is executed by the user selecting an arbitrary one from an inference engine selection switch (not shown) on the input panel 181. If an inference engine to be executed in the inference information creation apparatus 1 is set in advance, the set inference engine is automatically read and selected. If the user does not select an inference engine, a default inference engine is automatically set.

  Next, inference data creation processing (S3) for creating inference data based on the information measured from the sensor is executed according to the inference engine selected in S2. As described above, the processing contents in the inference data creation processing can be various depending on the inference engine. In this embodiment, the processing by the inference engine that creates inference data based on the measurement values from each sensor. An example is shown.

  As shown in FIG. 6, in the inference data creation process (S3), the measured values of the user's body temperature, sweat, and heart rate measured by the body temperature sensor 182, the sweat sensor 183, and the heart rate sensor 184 are respectively obtained. Obtained (S201). Next, a state variable which is a flag indicating a change in state of a measurement value from each sensor is cleared (S202). In the present embodiment, three bits of the second bit for the body temperature measured by the body temperature sensor 182, the first bit for the sweating measured by the sweat sensor 183, and the 0th bit for the heart rate measured by the heart rate sensor 184 are in a state. As a variable. Then, with reference to a reference value area (not shown) in the RAM 130, a change in the measured value from each sensor is determined based on the reference value of each sensor calculated in S107 (see FIG. 5).

  First, the body temperature measurement value acquired from the body temperature sensor 182 is compared with a reference value (body temperature threshold) related to body temperature (S203). The reference value (threshold value) is a value stored in a reference value area (not shown) of the RAM 130. As a result, if the measured body temperature is larger than the body temperature threshold (S203: YES), the second bit is set to “UP” (S204). On the other hand, when the measured body temperature value is not larger than the body temperature threshold value (S203: NO), the process proceeds to the next step (S205). Similarly, the perspiration measurement value acquired from the perspiration sensor 183 is compared with a reference value (perspiration threshold value) related to perspiration (S205). As a result, if the perspiration measurement value is greater than the perspiration threshold value (S205: YES), The first bit is set to “UP” (S206). On the other hand, when the perspiration measurement value is not larger than the perspiration threshold (S205: NO), the process proceeds to the next step (S207) as it is. In addition, the heart rate measurement value acquired from the heart rate sensor 184 is also compared with a reference value (heart rate threshold value) relating to the heart rate (S207). As a result, if the heart rate measurement value is greater than the heart rate threshold value (S207). : YES), 0th bit is set to “UP” (S208). On the other hand, when the heart rate measurement value is not larger than the heart rate threshold value (S207: NO), the process proceeds to the next step (S209) as it is.

  Thereafter, inference types and inference values corresponding to the patterns of the second bit, the first bit, and the 0th bit of the state variable are acquired from the inference definition table 13 (S209). As shown in FIG. 7, the inference definition table 13 includes an inference type 13a indicating the type of inference, a sensor state 13b indicating a change state of a measured value from each sensor, and an inference indicating the strength of the user's inference numerically. The value 13c is included as a data item. And the correspondence of each data item is defined in a table format. As described above, in this embodiment, inference data relating to the “impression” of the user is created, so the inference definition table 13 referred to also relates to “impression”. Therefore, the inference type 13a defines a type related to “impression”, and there are a plurality of types from “excitement” to “no impression (normal)” depending on the strength of the user's “impression”. Also, an inference value 13c that expresses the strength of the “feeling” as a numerical value is defined. For example, if the inference type 13a is “big excitement”, the inference value 13c is the maximum value “100”. Since the inference definition table 13 shown in FIG. 7 relates to “impression”, the inference value 13c is also displayed as the impression level (E). In S209, since the sensor state 13b is specified by the state variables set in S203 to S208, the inference type 13a and the inference value 13c corresponding to the sensor state 13b are acquired. Finally, inference data including the inference type and inference value acquired in S209 is created (S210).

  As described above, the inference engine selected in S2 is executed by the CPU 110, the inference data creation process (S3) is executed, and inference data is created. As described above, since a plurality of inference engines are provided and the user can select any inference engine, the inference data can be created by the optimum inference engine according to the use situation or use environment of the inference information creation apparatus 1. In addition, inference data about the user can be created more accurately.

  Returning to FIG. 4, the inference information output process (S11) for outputting the inference data created in the inference data creation process (S3) is executed. As shown in FIG. 8, in the inference information output process (S11), an inference engine ID is first acquired and added to the inference data created in the inference data creation process (S3) (S301). As described above, since the inference engine ID is stored in each inference engine stored in the HDD 140, the inference engine executed in S3 is referred to, and the inference engine ID is acquired and added to the inference data. Is done. Then, inference information is created based on the inference data and the inference engine ID (S302). As shown in FIG. 9, the inference information 10 includes at least an inference value 10a, an inference type 10b, and an inference engine ID 10c. The inference value 10a and the inference type 10b correspond to the inference value 13c and the inference type 13a included in the inference data, respectively. The inference information 10 created in S302 is stored in the inference information storage area 144 (see FIG. 3) of the HDD 140 (S303).

  Thereafter, returning to FIG. 4, it is determined whether or not a predetermined time has passed (S12). This predetermined time is a time set in advance in the time measuring device 190, and in S12, with reference to the time measuring device 190, the passage of the predetermined time is determined. The predetermined time set in the time measuring device 190 can be set arbitrarily by the user or the designer.

  If the predetermined time has not elapsed (S12: NO), the process returns to S12 to be in a waiting state, and this loop processing is repeated until the predetermined time has elapsed. On the other hand, when the predetermined time has elapsed (S12: YES), the process returns to S3, the inference data creation process (S3) is executed to create inference data, and the inference information 10 is output (S11). That is, in the inference information creation device 1, the process of outputting the latest inference information 10 is repeated every predetermined time. As a result, the inference information storage area 144 (see FIG. 3) of the HDD 140 stores a plurality of inference information 10 related to the user in time series.

  As described above, in the “inference information creation process”, since the inference value 10a and the inference type 10b are inferred based on the measurement values from the sensors, the inference information 10 that accurately infers the user's consciousness and emotion is created. be able to. Moreover, since the inference engine ID 10c indicating the origin of the inference engine is assigned to the inference information 10, the origin of the inference engine can be clarified to increase the reliability of the inference information 10.

  Next, the “inference information characteristic-specific processing” will be described with reference to FIGS. 10 to 15. FIG. 10 is a main flowchart of inference information characteristic-specific processing. FIG. 11 is a diagram illustrating a data configuration of the characteristic information table 15. FIG. 12 is a flowchart showing details of the characteristic-specific process A. FIG. 13 is a flowchart showing details of the characteristic-specific process B. FIG. 14 is a flowchart showing details of the characteristic-specific processing C. FIG. 15 is a flowchart showing details of the characteristic-specific processing D.

  In the following “inference information characteristic-specific processing”, as an example, a case will be described in which a user creates an earthquake distribution map, a report, or the like using the inference information for which the characteristic-specific processing has been executed.

  The main flowchart (FIG. 10) of the inference information characteristic-specific process is performed when the user instructs a process from the characteristic-specific process instruction switch (not shown) of the input panel 181 or every elapse of a predetermined time, or S303. When the inference information is stored in the inference information storage area 144, the processing is started. The execution timing of this inference information characteristic-specific processing can be arbitrarily set by the user or the designer. However, in this embodiment, when the user instructs from the characteristic-specific processing instruction switch of the input panel 181, The process shall be started.

  As shown in FIG. 10, in the inference information characteristic-specific processing, the inference information 10 to be processed by characteristic is read from the inference information storage area 144 (see FIG. 3) of the HDD 140 (S21). Here, all inference information 10 stored in the inference information storage area 144 may be read, or only a part may be read. Further, the inference information 10 to be read may be selected by the user. Next, the inference engine ID 10c of the inference information 10 to be processed is acquired (S22). As described above, the inference information 10 stores the inference engine ID 10c, which is identification information unique to the inference engine used to create each inference information (see FIG. 9), and this inference engine ID 10c is used as the inference information. 10 is obtained.

  Then, the characteristic information corresponding to the inference engine ID 10c is acquired with reference to the characteristic information table 15 (S23). As shown in FIG. 11, the characteristic information table 15 includes a reliability 15b, an update date 15c, and an inference type 15d as data items in addition to the inference engine ID 15a, and is defined in association with each other in a table format. The reliability 15b is data indicating the accuracy of the inference engine, and the higher this value, the more accurate inference is possible. The update date 15c indicates the latest update date of the inference engine. The newer the date, the more recently it is created or updated.

  The inference type 15d indicates the type of inference method of the inference engine. For example, the inference engine of the inference type “AA” refers to table information (LUT) such as the inference definition table 13 based on the measurement value from the sensor. This type of reasoning. The inference engine of the inference type “BB” is a type that infers by performing predetermined arithmetic processing on the measurement value from the sensor. The inference engine of the inference type “CC” is a type that processes and infers the measurement value from the sensor according to a predetermined procedure. The inference engine of the inference type “DD” is a hybrid inference type that performs inference by combining a plurality of methods from the above “AA”, “BB”, and “CC”. Therefore, in the present embodiment, the inference engine for executing the inference data creation process of FIG. 6 corresponds to the inference type “AA”.

  Since the characteristic information table 15 is a definition file relating to the properties of each inference engine, the latest definition file is acquired from an external storage medium or network by the user or automatically, and the characteristic information table 15 is stored in the characteristic information table 15. It shall be updated regularly with the latest definition file.

  In S23, the inference engine ID 15a, the reliability 15b, the update date 15c, and the inference type 15d are acquired as characteristic information using the inference engine ID 10c included in the inference information 10 as an index. For example, when the inference information 10 shown in FIG. 9 is a processing target, since the inference engine ID 10c is “ABC-0011”, the inference engine ID 15a “ABC-0011”, the reliability 15b “20”, and the update date 15c “03 / 12/12 "and inference type 15d" CC "are acquired as characteristic information.

  Next, it is determined based on the characteristic information regarding which data item the process for each characteristic should be executed (S24), and the process for each characteristic is selected based on this determination result. For example, if “characteristic information = inference engine ID”, the characteristic-specific processing A (S25) based on the inference engine ID 15a acquired in S23 is executed. Similarly, if “characteristic information = reliability”, the characteristic-specific process B (S26) based on the reliability 15b is executed, and if “characteristic information = update date”, the characteristic-specific process C based on the update date 15c (S27) ) Is executed, and if “characteristic information = inference type”, characteristic-specific processing D (S28) based on the inference type 15d is executed. Note that the definition information indicating which data item should be processed is set in the HDD 140 or the like in advance, but this setting can be arbitrarily set by the user or the designer and can be changed as appropriate. .

  As shown in FIG. 12, in the characteristic-specific process A (process based on the inference engine ID), it is determined whether or not the inference engine ID 15a matches the inference engine ID of the inference engine for earthquakes (S401). When the inference engine ID of the inference engine for earthquakes matches the inference engine ID 15a (S401: YES), the inference information 10 is an earthquake data file (not shown) provided in the characteristic information storage area 145 (see FIG. 3) of the HDD 140. (S402). On the other hand, when the inference engine ID of the earthquake inference engine and the inference engine ID 15a do not match (S401: NO), the inference information 10 is discarded and the process returns to FIG. As a result, only inference information 10 created by an inference engine suitable for various inferences regarding earthquakes can be saved in an earthquake data file, and a user can create an earthquake distribution map or report using this earthquake data file. Can do.

  As shown in FIG. 13, in the characteristic-specific process B (process by reliability), it is determined whether or not the reliability 15b is “80” or more (S411). When the reliability 15b is “80” or more (S411: YES), the inference information 10 is stored in the earthquake data file (S412) as in S402. When the reliability 15b is not “80” or higher (S411: NO), it is determined whether or not the reliability 15b is “60” or higher (S413). When the reliability 15b is “60” or more (S413: YES), the inference information 10 is saved in an earthquake auxiliary data file (not shown) provided in the characteristic-specific information storage area 145 (see FIG. 3) of the HDD 140. (S414). On the other hand, when the reliability 15b is not “60” or more (S413: NO), the inference information 10 is discarded and the process returns to FIG. Thereby, the reasoning information 10 created by the reasoning engine with high reliability can be stored in the earthquake data file, and the reasoning information 10 created by the reasoning engine with relatively high reliability can be saved in the earthquake auxiliary data file. Using this seismic data file and seismic auxiliary data file properly, earthquake distribution maps and reports can be created.

  As shown in FIG. 14, in the characteristic-specific process C (process by update date), it is determined whether or not the update date 15c is greater than or equal to a numerical value indicating the date and time three months ago (S421). When the update date 15c is newer than the date and time three months ago (S421: YES), the inference information 10 is stored in the earthquake data file (S422), as in S402. If the update date 15c is not newer than the date and time three months ago (S421: NO), it is determined whether or not the update date 15c is greater than or equal to a value indicating the date and time one year ago (new). When the update date 15c is newer than the date and time one year ago (S423: YES), the inference information 10 is stored in the earthquake auxiliary data file (S424), as in S414. On the other hand, if the update date 15c is not newer than the date one year ago (S423: NO), the inference information 10 is discarded and the processing returns to FIG. As a result, the inference information 10 created by the inference engine with the new update date can be saved in the earthquake data file, and the inference information 10 created by the inference engine with the relatively new update date can be saved in the earthquake auxiliary data file. Using this seismic data file and seismic auxiliary data file properly, earthquake distribution maps and reports can be created.

  As shown in FIG. 15, in the characteristic-specific process D (process based on the inference type), it is determined whether or not the inference type 15d matches “AA” (S431). When the inference type 15d is “AA” (S431: YES), the inference information 10 is stored in the earthquake data file (S432), as in S402. If the inference type 15d is not “AA” (S431: NO), it is determined whether or not the inference type 15d matches “BB” (S433). When the inference type 15d is “BB” (S433: YES), the inference information 10 is stored in the earthquake auxiliary data file (S434) as in S414. On the other hand, when the inference type 15d is not “BB” (S433: NO), the inference information 10 is discarded and the process returns to FIG. As a result, the inference information 10 created by the inference engine of the inference type “AA” can be saved in the earthquake data file, the inference information 10 created by the inference engine of the inference type “BB” can be saved in the earthquake auxiliary data file, etc. Users can categorize inference information into seismic data files and seismic auxiliary data files according to inference types, and create earthquake distribution maps and reports.

  As described above, in the characteristic-specific processing (S25, S26, S27, S28) shown in any of FIGS. 12 to 15, the characteristic-specific processing is executed based on the characteristic information, and the user can obtain an earthquake distribution map or report. For example, it is possible to obtain effective and convenient data.

  In the present embodiment, the case where each characteristic process (S25, S26, S27, S28) is executed for “earthquake” is described as an example, but these processes are limited to those for “earthquake”. Needless to say, this is not done. For example, each characteristic process may be performed according to the type of emotion, consciousness, situation, event, atmosphere, object, subject, and the like. The details of the processing in each characteristic process (S25, S26, S27, S28) are not limited to those described above. In addition to classifying the inference information 10 for each file, the inference information 10 is processed and corrected. Various processes such as combining them into a single document are possible, and the user or designer may set an optimal process arbitrarily.

  Returning to FIG. 10, it is determined whether or not the processing has been executed for all the inference information 10 to be processed (S29). If the processing has not been executed for all the inference information 10 to be processed (S29: NO), the processing returns to S21 because unprocessed inference information 10 remains. Therefore, the processes of S21 to S29 are repeated until there is no unprocessed inference information 10. On the other hand, when the process has been executed for all inference information 10 to be processed (S29: YES), this process ends.

  As described above, in the “inference information characteristic-specific processing”, processing according to the characteristics is executed, and effective and convenient data can be obtained for the subsequent use of the inference information 10, so that the use range of the inference information 10 is further expanded. be able to.

  Next, a second embodiment of the present invention will be described with reference to the drawings. The inference information management system according to the second embodiment includes an inference information creation device that is a small portable terminal device carried by a user and an inference information management device that is a fixed computer device via a network. Connected system.

  In the inference information management system of this embodiment, inference information created by each of a plurality of inference information creation devices is collected in the inference information management device via a network, and processing according to characteristics is performed on the inference information. The case where is executed is illustrated.

  First, the configuration of the inference information management system according to the second embodiment will be described with reference to FIGS. FIG. 16 is an overall configuration diagram of the inference information management system. FIG. 17 is a block diagram showing another configuration of the inference information creation apparatus 1. FIG. 18 is a block diagram illustrating a configuration of the inference information management device 2.

  As shown in FIG. 16, in the inference information management system according to the present embodiment, each of a plurality of inference information creation devices 1 and the inference information management device 2 are connected via a network 90. The network 90 only needs to be effectively connected so that data can be transmitted and received between terminals regardless of wired or wireless.

  As shown in FIG. 17, each of the inference information creation apparatuses 1 has basically the same configuration as that of the first embodiment (see FIG. 1). However, it differs in that a communication unit 170 for connecting to an external network 90 is provided. The communication unit 170 only needs to be able to be effectively connected to the external network 90 by wire or wireless. For example, in this embodiment, the communication unit 170 is a wireless LAN adapter for executing connection with the network 90 by a wireless LAN. . Further, the present embodiment is different from the first embodiment in that an inference engine, a characteristic information table 15, and a characteristic-specific processing instruction switch (not shown) that are configurations related to “inference information characteristic-specific processing” are not provided.

  As shown in FIG. 18, the inference information management apparatus 2 includes a CPU 210, a ROM 220, a RAM 230, an HDD 240, a display control unit 260 to which a display 261 is connected, a voice control unit 270 to which a microphone 271 and a speaker 272 are connected, and a mouse 281. And an input detection unit 280 to which a keyboard 282 is connected is provided via a bus 215. Such a configuration of the inference information management apparatus 2 is known as a configuration of a normal computer device, and thus details thereof are omitted. The inference information management device 2 is provided with a communication interface 291 for effectively connecting to the external network 90 by wire or wireless. The communication interface 291 is only required to be able to be effectively connected to the network 90. For example, in this embodiment, the communication interface 291 is a LAN card for cable connection to a wired LAN.

  In addition, the inference information management device 2 includes an inference engine, a characteristic information table 15, and a characteristic-specific processing instruction switch (not shown) that are components related to “inference information characteristic-specific processing”. Specifically, a plurality of inference engines are stored in the program storage area (not shown) of the HDD 240, and the characteristic information table 15 is stored in the information storage area (not shown). The HDD 240 is provided with a characteristic-specific information storage area (not shown), and the mouse 181 or the keyboard 282 functions as a characteristic-specific processing instruction switch (not shown).

  Hereinafter, the flow of processing in the inference information management system of the present invention will be described with reference to FIGS. 19 and 20. FIG. 19 is a flowchart showing other details of the inference information output process (S11) in the inference information creation apparatus 1. FIG. 20 is another main flowchart of the process for each inference information characteristic in the inference information management apparatus 2.

  First, processing in the inference information creation apparatus 1 will be described. In the inference information creation device 1, as in the first embodiment, the “inference information creation process” (FIG. 4) is executed, but the inference information output process (FIG. 8) in S 11 is different.

  As shown in FIG. 19, in the inference information output process (S11), as in FIG. 8, an inference engine ID is acquired and added to the inference data (S301). Based on the inference data and the inference engine ID, Inference information is created (S302). Then, the inference information created in S302 is transmitted from the communication unit 170 to the inference information management apparatus 2 via the network 90 (S304). The inference information transmitted to the inference information management device 2 is the same as the inference information 10 shown in FIG.

  Note that the execution timing of the transmission process in S304 is not limited to the creation of the inference information 10, but the inference information 10 created in S302 is stored in the inference information storage area 144 (see FIG. 3) of the HDD 140 for a predetermined period. Every time or when there is an instruction from the user, the transmission process of S304 may be executed.

  Next, processing in the inference information management device 2 will be described. In the inference information management device 2, the “inference information characteristic-specific processing” (FIG. 10) in the first embodiment is executed, but the processing is executed based on the inference information 10 transmitted from the inference information creation device 1. It is different in point. In the present embodiment, when the inference information management device 2 receives the inference information 10 transmitted from the inference information creation device 1, this process is started.

  As shown in FIG. 20, in the inference information characteristic-specific processing, the inference information 10 transmitted via the network 90 is received by the communication interface 291 and stored in an inference information storage area (not shown) of the HDD 240. (S20). The subsequent processing is the same as the processing (S21 to S29) shown in FIG.

  Note that the execution timings of S21 to S29 are not limited to when the inference information 10 is received in S20, and after the inference information 10 received in S20 is stored in the inference information storage area of the HDD 240, every predetermined period of time or user When there is an instruction from, the processes of S21 to S29 may be executed.

  As described above, according to the inference information management system of the second embodiment, the inference information 10 created by the inference information creation device 1 is collected and managed by the inference information management device 2, and the inference information management device In 2, the process according to the characteristic is executed. Therefore, the inference information creation apparatus 1 that executes the “inference information creation process” and the inference information management apparatus 2 that executes the “inference information characteristic-specific process” can be configured independently of each other, and the inference information can be more flexibly provided. A management system can be configured. Further, the origin of the inference engine can be clarified, the reliability of the inference information 10 can be improved, and the range of use can be expanded.

  By the way, in the first and second embodiments, the input detection unit 180 (FIG. 1) that detects the input of the measurement value from the body temperature sensor 182, the sweat sensor 183, and the heart rate sensor 184 is the “measurement” of the present invention. It corresponds to “value acquisition means”. Further, the CPU 110 that executes the inference engine corresponding to the inference data creation process (S3) shown in FIG. 6 corresponds to the “inference data creation means” of the present invention. In the inference information output process (S11) shown in FIG. 8 or FIG. 19, the CPU 110 that executes S301 corresponds to the “identification information adding means” of the present invention, and the CPU 110 that executes S302, S303, and S304 The communication unit 170 (FIG. 17) corresponds to the “first communication interface unit” of the present invention. The inference engine identification ID 10c corresponds to “identification information” of the present invention.

  Further, in the “inference information characteristic-specific processing” (FIG. 10), the CPU 110 that executes S23 corresponds to the “characteristic information acquisition unit” of the present invention, and the CPU 110 that executes S24 performs the “processing procedure selection unit” of the present invention. The CPU 110 that executes the characteristic-specific processing (S25, S26, S27, S28) corresponds to the “inference information processing means” of the present invention. The characteristic-specific processing (S25, S26, S27, S28) corresponds to the “processing procedure” of the present invention. The CPU 110 that executes S20 corresponds to the “inference information acquisition unit” of the present invention, and the communication interface 291 (FIG. 18) corresponds to the “second communication interface unit” of the present invention. The inference information storage area 144 provided in the HDD 140 or the inference information storage area (not shown) provided in the HDD 240 corresponds to the “inference information storage means” of the present invention.

  The present invention is not limited to the first and second embodiments described in detail above, and it goes without saying that various modifications are possible.

  For example, in the above embodiment, the case where the inference information about the user is related to “impression” has been described as an example, but the inference information is not limited to the user's emotion and consciousness, but also the atmosphere, the importance level, etc. It may be information about an abstract concept (also referred to as a context) that indicates the context of an event or the context of a situation, and cannot be grasped only by facts and evidence. Therefore, inference information may be created for “sadness”, “anger”, “fun”, “liveness”, “busy”, and the like. Then, an inference definition table 13 corresponding to the inference contents of each inference information may be set. For example, in order to create inference information based on the user's “sadness”, the inference definition table 13 corresponding to “sadness” may be set.

  In addition, in the inference definition table 13, a table for arbitrary inference contents may be set in advance by the user or the designer, and a plurality of tables corresponding to each of a plurality of inference contents are included in the inference definition table 13. It may be set in advance, and an optimum table may be automatically selected in the inference data creation process (FIG. 6).

  As characteristic information of the inference engine, the characteristic information table 15 defines an inference engine ID 15a, a reliability 15b, an update date 15c, and an inference type 15d. However, the characteristic information is not limited to these items. Absent. For example, various characteristics such as an inference engine manufacturer, version information, and inference contents can be arbitrarily defined by a user or a designer. Further, the inference engine may be implemented as software (program) or hardware such as an electric circuit or a device device.

  In addition, as a processing procedure to be executed for the pre-stored inference information, if it is determined that the characteristic information is unsuitable for the subsequent processing in terms of reliability or update date, it is again determined by another inference engine. Inference information may be acquired, or correction by an appropriate correction value may be performed to convert the inference information into an appropriate value.

  In addition, in the inference engine initialization process (FIG. 5), the sampling value is measured and the average value is calculated as the reference value. However, the time series data of the sampling value is obtained, and based on the characteristics of the transition. The reference value may be calculated. Further, the reference value may be calculated by excluding abnormal sampling values. Further, in the comparison process (S203, S205, S207) in the inference data creation process (FIG. 6), a change threshold value ε is provided for each sensor, and a correction value obtained by correcting the threshold value with the change threshold value ε is obtained from each sensor. Comparison with the measured value may be executed. For example, about 5% of the threshold is set with the change threshold ε as an allowable error range. Further, in the present embodiment, the state change is determined by comparing the measured value from each sensor with a threshold value, but a predetermined reference value is subtracted from the measured value from each sensor to obtain an increment value, The state change may be determined by comparing whether the increment value is larger or smaller than a threshold value.

  Needless to say, the measured values from each sensor are not limited to body temperature, sweating and heart rate. For example, the vibration, brain wave, respiration, acceleration, inclination, biorhythm, etc. of the user may be measured from the user. Further, each sensor (body temperature sensor 182, sweat sensor 183, heart rate sensor 184) and input panel 181 do not have to be configured integrally with the inference information creation device 1, and are input via an interface such as a USB or network. It is only necessary to remotely connect to the detection unit 180 and effectively acquire measurement values and input information.

  Similarly, in the inference information management device 2, the display 261, microphone 271, speaker 272, mouse 281, and keyboard 282 are not necessarily required configurations, so that an external display device, microphone, Remote control may be performed by remotely connecting to a speaker or the like.

  A plurality of inference information management devices 2 may be provided in the inference information management system. The inference information creation device 1 and the inference information management device 2 may be configured as an integrated device. Conversely, the inference information creation device 1 may be one.

  The inference information creation device, the inference information management system, and the inference information creation program of the present invention can be applied to a computer device for inferring the user's consciousness and emotion.

It is a block diagram which shows the structure of the inference information creation apparatus. 3 is a conceptual diagram illustrating a configuration of a storage area of a RAM 130 of the inference information creation device 1. FIG. 4 is a conceptual diagram illustrating a configuration of a storage area of an HDD 140 of the inference information creation device 1. FIG. It is a main flowchart of inference information creation processing. It is a flowchart which shows the detail of an inference engine initialization process (S1). It is a flowchart which shows the detail of an inference data creation process (S3). It is a figure which shows the data structure of the inference definition table. It is a flowchart which shows the detail of an inference information output process (S11). It is a figure which shows the data structure of the inference information 10. It is a main flowchart of a process according to inference information characteristics. 6 is a diagram illustrating a data configuration of a characteristic information table 15. FIG. It is a flowchart which shows the detail of the process A classified by characteristic. It is a flowchart which shows the detail of the process B by characteristic. It is a flowchart which shows the detail of the process C classified by characteristic. It is a flowchart which shows the detail of the process D classified by characteristic. It is a whole block diagram of an inference information management system. It is a block diagram which shows the other structure of the inference information creation apparatus. 3 is a block diagram showing a configuration of an inference information management device 2. FIG. It is a flowchart which shows the other details of an inference information output process (S11) in the inference information creation apparatus 1. It is another main flowchart in the inference information management device 2 according to inference information characteristic processing.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Inference information creation apparatus 2 Inference information management apparatus 10 Inference information 13 Inference definition table 15 Characteristic information table 90 Network 110 CPU
115 bus 120 ROM
130 RAM
140 HDD
170 Communication Unit 180 Input Detection Unit 181 Input Panel 182 Body Temperature Sensor 183 Sweating Sensor 184 Heart Rate Sensor 190 Timing Device 210 CPU
215 bus 220 ROM
230 RAM
240 HDD
260 Display Control Unit 261 Display 270 Audio Control Unit 271 Microphone 272 Speaker 280 Input Detection Unit 281 Mouse 282 Keyboard 291 Communication Interface

Claims (9)

Measurement value acquisition means for acquiring measurement values from at least one sensor;
Based on the measurement value acquired by the measurement value acquisition means, inference data creation means for creating inference data that is an index value different from the measurement value;
Identification information adding means for adding identification information unique to the inference data creating means to the inference data;
And an inference information output unit that outputs inference information including the inference data to which the identification information is added. A characteristic information table for storing the identification information of the inference data creation means and the characteristic information indicating the characteristics of the inference data creation means in association with each other;
The characteristic information acquisition means for acquiring the characteristic information corresponding to the identification information included in the inference information output by the inference information output means from the characteristic information table. Inference information creation device.   3. The inference information creation apparatus according to claim 2, wherein the characteristic information includes at least one of reliability, latest update date, and inference type of the inference data creation means. A processing procedure selection unit that includes at least one processing procedure to be executed for the inference information and selects one processing procedure from a plurality of the processing procedures based on the characteristic information acquired by the characteristic information acquisition unit. When,
The inference information processing means for processing the inference information output by the inference information output means on the basis of the processing procedure selected by the processing procedure selection means. Inference information creation device. An inference information creation device for creating inference information of a user based on a measurement value acquired from at least one sensor, and an inference information management device for managing the inference information created by the inference information creation device; Is an inference information management system connected via a network,
The inference information creation device includes:
A measurement value acquisition means for acquiring the measurement value from the sensor;
Based on the measurement value acquired by the measurement value acquisition means, inference data creation means for creating inference data that is an index value different from the measurement value;
Identification information addition means for adding identification information unique to the inference data creation means to the inference data;
Inference information output means for outputting the inference information including the inference data to which the identification information is added,
The inference information management device includes:
Inference information acquisition means for acquiring the inference information output from the inference information creation device via a network;
Inference information storage means for storing the inference information acquired by the inference information acquisition means;
A characteristic information table for storing the identification information of the inference data creation means and the characteristic information indicating the characteristics of the inference data creation means in association with each other;
An inference information management system comprising: characteristic information acquisition means for acquiring, from the characteristic information table, the characteristic information corresponding to the identification information included in the inference information output by the inference information output means.   6. The inference information management system according to claim 5, wherein the characteristic information includes at least one of reliability, latest update date, and inference type of the inference data creation means. The inference information management device includes:
A processing procedure selection unit that includes at least one processing procedure to be executed for the inference information, and selects one processing procedure from a plurality of the processing procedures based on the characteristic information acquired by the characteristic information acquisition unit. When,
The inference information processing means for processing the inference information output by the inference information output means on the basis of the processing procedure selected by the processing procedure selection means. Inference information management system. The inference information output means includes a first communication interface means for executing data transmission / reception with the inference information management device wirelessly or by wire.
8. The inference according to claim 5, wherein the inference information acquisition unit includes a second communication interface unit for executing data transmission / reception with the inference information creation device wirelessly or by wire. Information management system. Computer
Measurement value acquisition means for acquiring measurement values from at least one sensor;
Inference data creation means for creating inference data that is an index value different from the measurement value based on the measurement value acquired by the measurement value acquisition means;
Identification information adding means for adding identification information unique to the inference data creating means to the inference data;
Inference information output means for outputting inference information including the inference data to which the identification information is added;
Inference information creation program to function as

Images