JP2009181478A - Condition estimating device and condition estimating program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a condition estimating device and a condition estimating program, which can estimate user conditions without using a sensor or a tag. <P>SOLUTION: A user history log storage part 12 of a mobile communication terminal 100 stores a user history log which is a log of past actions of a user. Based on the user history log stored in the user history log storage part 12, a condition index calculating part 19 calculates condition indexes such as a "sleep index", a "labor index", a "vigor index", a "mannerism index", a "social dropout index", and a "tension index", which are indicators for estimating user conditions. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a condition estimation device and a condition estimation program for estimating a condition representing the state of mind and body of a user.

  2. Description of the Related Art Conventionally, various services are provided to a user based on the position and time at which the user is present. For example, Patent Document 1 discloses that based on the position of the user terminal acquired from the base station and the current time, the user is near his home in the evening, near his work place in the morning or noon, or in the evening or at night Predict the situation of the user such as getting on the bus and moving to the vicinity of the house, and according to the prediction result, "Is this your home?" The station is Kyodai Station near the house. Please get off to the right. ”A system that displays a message such as“ on the user terminal ”is disclosed. However, even if the situation is estimated based only on the current position and time of the user, the accuracy is low. Therefore, the message to be displayed based on the estimation result with low accuracy must be in the question format as described above, and an appropriate message cannot be displayed to heal the user.

Conventionally, in addition to situation estimation as described above, by using sensors and tags, it is possible to estimate the user's condition such as whether the user is fine, not depressed, or withdrawn. Service is also provided to users.
JP 2006-65870 A

However, it is difficult to easily estimate the condition without using a sensor or a tag.
The present invention has been made to solve the above-described problems of the prior art, and is a condition estimation device and a condition estimation program that can easily estimate a user's condition without using a sensor or a tag. Is to provide.

In order to solve the above-mentioned problem, the invention according to claim 1 of the present invention stores a user history log, which is a history of a user's past actions, in a condition estimation device for estimating a condition representing the state of mind and body of the user. User history log storage means, and condition index calculation means for calculating a condition index serving as an index for estimating the condition of the user based on the user history log stored in the user history log storage means. A condition estimation device is provided.
According to the present invention, the condition estimation device calculates a condition index that serves as an index for estimating the user's condition based on the user history log. Therefore, the condition estimation device calculates the condition of the user without using a sensor or a tag. It is possible to easily estimate the condition index as an index for estimation.

  The invention according to claim 2 is the condition estimation device according to claim 1, wherein the user's dependence on life cycle information representing a daily life cycle of the user and position information representing a place where the user is located. Model definition means for storing a situation transition model in which the order of situation transitions representing the state of behavior is defined according to weekdays, position information acquisition means for acquiring position information indicating the current location of the user, and current time Time information acquisition means for acquiring time information to represent, the situation transition model, the position information acquired by the position information acquisition means, the time information acquired by the time information acquisition means, and the situation before the user transition A situation estimator that estimates the user's current situation based on DOO further wherein the user log storage unit, the current situation estimated by the situation estimating means, and a time when the transition to the current situation, and to store, as the user log.

  According to the present invention, the condition estimation device stores the current situation estimated by the situation estimation unit and the time of transition to the current situation as the user history log. A highly accurate user history log can be stored based on the estimation, and a condition index serving as an index for estimating the condition of the user can be estimated with high accuracy based on the highly accurate user history log. It becomes possible.

  According to a third aspect of the present invention, in the condition estimation device according to the first or second aspect, the condition index includes a sleep index serving as an index as to whether or not the user's sleep is sufficient, and the condition The index calculation means calculates an average value of the user's past sleep time based on the user history log, and if the latest sleep time of the user is longer than the average value, the sleep index is calculated for the user. The sleep index is updated to a value indicating that sleep is more sufficient, and when the average value is shorter, the sleep index is updated to a value indicating that the user's sleep is insufficient.

  According to the present invention, the condition index calculation means calculates an average value of the user's past sleep time based on the user history log, and when the latest sleep time of the user is longer than the average value. Updates the sleep index to a value indicating that the user's sleep is more sufficient, and if shorter than the average value, updates the sleep index to a value indicating that the user's sleep is less sufficient Therefore, the sleep index can be easily calculated without using a sensor or a tag, and it can be easily estimated whether or not the user's sleep is sufficient based on the sleep index.

  The invention according to claim 4 is the condition estimation device according to any one of claims 1 to 3, wherein the condition index includes a labor index serving as an index as to whether or not the user is working too much, The condition index calculating means calculates an average value of the past working hours of the user based on the user history log, and if the latest working time of the user is longer than the average value, the labor index is It is updated to a value indicating that the user's working hours are longer, and when the user's working time is shorter than the average value, the labor index is updated to a value indicating that the user's working hours are shorter.

  According to the present invention, the condition index calculating means calculates an average value of the past working hours of the user based on the user history log, and when the latest working time of the user is longer than the average value. To update the labor index to a value representing that the user's working hours are longer, and to update the labor index to a value representing that the user's working hours are shorter if shorter than the average value, The labor index can be easily calculated without using a sensor or tag, and it can be easily estimated whether or not the user is working too much based on the labor index.

According to a fifth aspect of the present invention, in the condition estimation device according to the fourth aspect, the condition index includes a spirit index indicating a degree of the user's energy, and the condition index calculation means includes the sleep index and The energy index is calculated based on the labor index.
According to the present invention, the condition index calculating means calculates the energy index based on the sleep index and the labor index, and therefore, based on the calculated energy index without using a sensor or a tag. The degree of the user's spirit can be easily estimated.

The invention according to claim 6 is the condition estimation device according to any one of claims 1 to 5, wherein the condition index includes a tension index indicating a degree of uplift of the user's mood, and the user The response to the question is classified as a positive factor response that is a response when the user's mood is uplifted or a negative factor response that is a response when the user's mood is depressed Response classification means for determining the condition index, and the condition index calculation means updates the tension index based on a determination result by the response classification means.
According to the present invention, since the condition index calculating means updates the tension index based on the determination result by the response classification means, the tension index can be easily calculated without using a sensor or a tag. The user's mood state can be easily estimated based on the tension index.

  The invention according to claim 7 is the condition estimation device according to any one of claims 2 to 6, wherein the condition index serves as an index as to whether or not there is a change in the daily life pattern of the user. A manner index is included, and the condition index calculation means calculates, based on the user history log, the time during which the user's situation is returning from work to return home for each day, and the calculated home return for each day When the difference in time is within the first predetermined time, the manner index is updated to a value indicating that the change in the lifestyle pattern of the user is smaller, and the difference in time during the return home for each day is When the first predetermined time is exceeded, the manner index is updated to a value indicating that the change in the life pattern of the user is larger.

  According to the present invention, the condition index calculating means calculates a time of the user returning home for each day based on the user history log, and the manner index according to the calculated difference in the time of returning home for each day. Therefore, it is possible to easily calculate the manner index without using a sensor or a tag, and it is possible to easily determine the lack of change in the life pattern of the user based on the manner index.

  The invention according to claim 8 is the condition estimation device according to claim 7, wherein the condition index calculation means is in a state where the user's situation is between work start and work start based on the user history log. Is calculated for each day, and when the calculated difference in working hours for each day is within a second predetermined time, the manner index is a value indicating that the change in the life pattern of the user is smaller Updating the manner index to a value indicating that the change in the user's life pattern is larger when the difference in the working hours for each day exceeds the second predetermined time. It is characterized by.

  According to the present invention, the condition index calculating means calculates a time during which the user is working for each day based on the user history log, and the manner index according to the calculated difference in the time during which the user is working. Therefore, the manner index can be easily calculated without using a sensor or a tag, and the lack of change in the user's life pattern can be easily estimated based on the manner index.

  The invention according to claim 9 is the condition estimation device according to any one of claims 2 to 8, and represents the correspondence relationship between the area representing the place where the user stays according to the purpose and the position information. Area definition means for storing area definition information, wherein the condition index includes a withdrawal index indicating a withdrawal tendency in which the user avoids participation in society, and the condition index calculation means includes the user history Based on the log and the area definition information, the user's situation on a holiday is at home, the energy index indicates that the degree of energy of the user is relatively low, and the user has a third If it is determined that the user has stayed in the area where the user resides for a predetermined time or more, the withdrawal index is set to the withdrawal index. Direction, characterized in that the updating to a value indicating that higher.

  According to the present invention, the condition index calculation means indicates that the user's situation on a holiday is at home, and the energy index indicates that the degree of energy of the user is relatively low, and In order to update the withdrawal index to a value indicating that the withdrawal tendency is higher when it is determined that the user stays in the area where the user resides for a third predetermined time or longer, a sensor or tag is Without use, the withdrawal index can be easily calculated based on the situation history stored in the user history log, and the user's withdrawal tendency can be easily calculated based on the withdrawal index. Can be estimated.

According to a tenth aspect of the present invention, in the condition estimation device according to the ninth aspect, when the condition index calculating means determines that the user has moved out of the area where the user resides, the withdrawal index is determined as the user. It is characterized by updating to a value indicating that the tendency of withdrawal is lower.
According to the present invention, when it is determined that the user has moved out of the residential area based on the user history log, the withdrawal index is updated to a value indicating that the withdrawal tendency of the user is lower. Therefore, the withdrawal index can be easily calculated without using a sensor or a tag, and the user's tendency to withdraw can be easily estimated based on the withdrawal index.

The invention according to claim 11 is the condition estimation device according to any one of claims 1 to 10, wherein a line for presenting to the user based on the condition index calculated by the condition index calculation means is provided. It further comprises a line deciding means for deciding.
According to the present invention, the serif determining means determines a serif to be presented to the user based on the condition index calculated by the condition index calculating means, so that an appropriate serif according to the user's condition is determined. It can be presented to the user, and healing can be provided by giving the user the impression that there is a person watching.

  The invention according to claim 12 is based on a user history log storage unit that stores a user history log that is a history of past actions of a user, and a user history log stored in the user history log storage unit. And a condition estimation program for functioning as a condition index calculating means for calculating a condition index serving as an index for estimating a condition representing the state of mind and body of the user.

  According to the present invention, since a condition index serving as an index for estimating the user's condition based on the user history log is calculated, it becomes an index for estimating the user's condition without using a sensor or a tag. It is possible to easily estimate the condition index.

Hereinafter, an embodiment when the condition estimation device according to the present invention is realized by a mobile communication terminal will be described with reference to the drawings. In the drawings referred to in the following description, the same parts as those in the other drawings are denoted by the same reference numerals.
(overall structure)
FIG. 1 is a diagram showing an overall configuration of an agent system according to an embodiment of the present invention. As shown in the figure, the agent system is based on a foldable mobile communication terminal 100 carried by a user, a server 200 that manages an application program for downloading to the mobile communication terminal 100, and a GPS (Global Positioning System). And a position positioning device 300 that measures the position of the mobile communication terminal 100.

(Hardware configuration of mobile communication terminal)
FIG. 2 is a block diagram illustrating a hardware configuration of the mobile communication terminal 100. As shown in the figure, the mobile communication terminal 100 displays a CPU 101 that controls the whole, a storage device 102 that stores software such as programs and data, a communication interface 103 for performing wireless communication, and various screens. A display device 104, an input device 105 that outputs an input signal corresponding to an operation such as pressing a key or opening / closing a terminal, a sound output device 106 that outputs a sound such as a sound or a sound effect, and a user's voice are input. A voice input device 107 such as a microphone, an internal clock 108 for measuring the date and time, and a GPS receiver 109 for receiving GPS signals from the position positioning device 300 are provided.
The storage device 102 includes a ROM 121, a RAM 122, and a nonvolatile memory 123. The ROM 121 stores an operating system, software for executing an application program compliant with the Java (registered trademark) specification, and the like. The RAM 122 temporarily stores various programs and data.

  The nonvolatile memory 123 is provided with an application storage area 123a and a scratch pad 123b. The application storage area 123a stores application programs downloaded from the server 200. Not only downloading but also the application program can be stored in the application storage area 123a from the time of manufacture of the mobile communication terminal 100. The scratch pad 123b stores data downloaded from the server 200 and data created by starting an application program or executing an application program. In the present embodiment, the scratch pad 123b stores calendar data for determining the day of the week, sound data for output from the sound output device 106, image data for display on the display device 104, and the like. The mobile communication terminal 100 may include an external storage device other than the scratch pad 123b as a memory for storing data.

(Functional configuration of mobile communication terminal)
Next, the functional configuration of the mobile communication terminal 100 will be described with reference to FIG. The functional configuration shown in the figure is realized by the CPU 101 of the mobile communication terminal 100 executing processes according to programs stored in the ROM 121, the application storage area 123a, and the like.
(Area definition part)
The area definition unit 11 defines the correspondence between the area representing the place where the user stays according to the purpose and the position information, and stores it as area definition information. In this embodiment, the area includes a home area (hereinafter referred to as “HA”), a temporary home area (hereinafter referred to as “temporary HA”), a work area (hereinafter referred to as “WA”), a temporary work area (hereinafter referred to as “temporary”). WA ”,“ favorite area ”, and“ specific area ”. Here, “HA” is a main place of residence of a user such as a home or a home. “Temporary HA” is a place where a user rarely goes to bed on a trip, business trip or the like, such as a friend's house or a hotel. “WA” is a user's main work place. “Temporary WA” is a place where a user rarely works for business, meetings, etc., such as a customer or event venue. The “favorite area” is an area set as a favorite place of the user. The “specific area” is an area set in order for the content provider to provide a service to the user in the area.

The area definition unit 11 defines an area name based on an input by the user from the input device 105. The area definition unit 11 stores the input area name in the scratch pad 123b in association with the position information of the mobile communication terminal 100.
In addition, the area definition unit 11 notifies the user of a question when the user stays in an undefined area for a predetermined time (for example, 1 hour) or more, and based on the answer to the question, the undefined area Define In addition, as a method of notifying a user of a question, a method of displaying the question content in characters on the display device 104 or a method of outputting the question content from the sound output device 106 by voice can be considered. In addition, as a method for a user to answer a question, a method of inputting answer contents from the input device 105 or a method of inputting voice from the voice input device 107 can be considered.

(User history log storage)
The user history log storage unit 12 stores a user history log, which is a history of past actions of the user, in the scratch pad 123b. FIG. 4 shows an example of the data structure of the user history log. As shown in the figure, the user history log includes date and time, weekday / holiday classification, location information, area type, situation, estimated wake-up time / estimated bedtime, displayed dialogue ID, answer to question, and terminal open / close Terminal operation logs such as time and key operation time are included. These user history logs are displayed on the display device 104 or the sound output device 106 when, for example, a situation estimation unit 17 described later estimates a situation, or when a model definition unit 13 described later calculates an estimated wake-up time or an estimated bedtime. It is stored when the speech is output, when the user answers a question, when positioning is performed by GPS, when the user operates the mobile communication terminal 100, or the like.

(Model definition part)
The model definition unit 13 defines a situation transition model and stores the defined situation transition model in the scratch pad 123b. Here, the “situation transition model” is a model that defines the transition order of the user's situation depending on the life cycle information representing the daily life cycle of the user and the position information representing the place where the user is present.
In addition, “life cycle information” refers to information representing a daily life cycle that is different for each user, such as the user's wake-up time, bedtime, start time, end time, wake-up time, and bedtime. The model definition unit 13 includes user-set time information indicating times such as wake-up time, bedtime, start time, and end time input by the user, and “wake-up time zone”, “sleep time”, and “wake-up time”. Life cycle information is defined by associating life cycle classification items such as “sleeping time”, “starting time”, and “working time”.

Further, the “situation” represents the state of the user's action. In the present embodiment, situations can be classified into, for example, point situations (PS), line situations (LS), and special situations (SS).
The “point situation (PS)” is a situation that is estimated based on the life cycle information of the user and the area where the user is located, and is the starting point of the action. Point situations include getting up, starting work, starting work, ending work, returning home, going to bed, going out, and the like. A point situation automatically transitions to the associated line situation.

A “line situation (LS)” is a situation that connects two point situations, and there are during preparation, going to work, working, overtime, going home, going home, sleeping, going out, and the like.
A “special situation (SS)” is a situation that is neither a point situation nor a line situation, and in this embodiment, there is a situation called “being lost”. This is a situation when the current user situation is unknown, such as at the first activation or before the first acquisition of position information.

  FIG. 5 shows an example of the situation transition model. In the figure, situations such as “sleeping”, “going out”, “getting up”, “medium condition” indicate life cycle information such as “before bedtime”, “before worktime”, and “HA”, The state of transition according to areas such as “WA” and “EX” is conceptually shown. Note that “EX” in the figure refers to an undefined area (except area) other than the areas defined in “HA”, “WA”, and the like.

FIG. 6 shows an example when the situation transition model shown in FIG. 5 is expressed in a table format. In the table shown in the figure, the situation before the user transition, the area (HA, provisional HA, WA, provisional WA, EX), life cycle information (wake-up time zone, start time, work time, etc.), current Is associated with the situation.
Further, in the columns (a) and (b) shown in FIG. 6, a situation transition order independent of position information (position-independent situation transition model) is defined. Specifically, in the column (a), the transition order of situations when the time during which user position information cannot be acquired is less than a certain time (for example, 2 hours) is defined. When the time during which the user's position information cannot be acquired is less than a predetermined time, the column (a) is referred to, and the current situation is determined depending only on the situation before the transition. On the other hand, in the column (b), the transition order of situations when the time during which the user's position information cannot be acquired continues for a certain time or more is defined. When the time during which the user's position information cannot be acquired continues for a certain time or longer, the column (b) is referred to, and the user's current situation is always estimated to be “lost” indicating that the situation is unknown. Note that the case where the position information cannot be acquired may be a case where the mobile communication terminal 100 cannot communicate with the position positioning device 300 because the mobile communication terminal 100 is out of the service area. Further, the time during which the user position information cannot be acquired can be calculated by referring to the user history log, for example. In the columns (a) and (b) shown in FIG. 6, the situation transition order that does not depend on the life cycle information in addition to the position information is defined, but depends on the life cycle information and does not depend only on the position information. A situation transition order may be defined.

  FIG. 7 shows a general situation transition pattern on weekdays, and FIG. 8 shows a general situation transition pattern on holidays. As shown in these figures, since the situation transition pattern generally differs between weekdays and holidays, the model definition unit 13 defines two types of situation transition models for weekdays and holidays, Are stored in association with each other.

(Correction of life cycle information)
The life cycle information is defined based on user set time information input by the user. The life cycle represented by the life cycle information may be different from the actual life cycle due to a change in the user's work style or the like. For this reason, the model definition unit 13 has a function of correcting life cycle information based on the user history log stored in the user history log storage unit 12. Specifically, for example, the model definition unit 13 uses the average value of the past estimated bedtime of the user and the average value of the estimated previous wake-up time, and the bedtime, wake-up time, and bedtime included in the life cycle information. Correct the belt, wake-up time, etc.

The average value μs and the standard deviation σ of the estimated bedtime are calculated by the following equations.
Average value μs = Σxi / N (N: total number of past estimated bedtime data, xi: data value)
Standard deviation σ = (Σ (xi−μs) ² / (N−1)) ^1/2
Similarly, the average value μw and the standard deviation σ of the estimated wake-up time are calculated by the following equations.
Average value μw = Σxi / N (N: total number of past estimated wake-up times, xi: data value)
Standard deviation σ = (Σ (xi−μw) ² / (N−1)) ^1/2
The model definition unit 13 estimates the latest bedtime and wake-up time at a predetermined timing every day. FIG. 9 is an example of a graph showing the number of times that each time was estimated as a bedtime / wake-up time in the past.

  Here, a method for estimating the latest bedtime will be described in detail. The model definition unit 13 refers to the user history log at a predetermined timing every day. For example, in the HA or temporary HA, the time when the operation such as the key operation and the opening / closing operation of the mobile communication terminal 100 last occurred (hereinafter referred to as “the first number”). 1) is used to estimate the latest bedtime. The model definition unit 13 estimates the bedtime by dividing the calculation logic into three types as shown in FIG. 9 according to the time zone when the first predetermined trigger occurs.

  Logic 1: A bedtime estimation logic when a first predetermined trigger occurs in a time zone before (μs−σ). Since there are few history logs of estimated bedtime in this time zone, the model definition unit 13 inquires the user about the latest bedtime of the user when a wake-up situation occurs in order to increase the certainty of estimation of the bedtime. . If there is an answer, the time of the answer is estimated as the bedtime. When there is no answer or when a wake-up situation has not occurred, the bedtime set by the user set time information is estimated as the latest bedtime.

  Logic 2: Bedtime estimation logic when a first predetermined trigger occurs in the bedtime (μs ± σ time zone) including the average value of the estimated bedtime of the user. However, when the number of past estimated bedtime data is 10 or less, 1 hour before and after the bedtime set by the user set time information is set as the bedtime. When the first predetermined trigger occurs during this bedtime, that time is set as the bedtime.

  Logic3: Bedtime estimation logic when the first predetermined trigger occurs in the time zone after (μs + σ). Since there are few history logs of the estimated bedtime in this time zone, the model definition unit 13 inquires of the user about the latest bedtime when the wake-up situation occurs in order to increase the certainty of the estimation of the bedtime. If there is an answer, the time of the answer is estimated as the bedtime. When there is no answer, or when a wake-up situation has not occurred, the bedtime set by the user-set time information is estimated as the bedtime.

  Next, a method for estimating the wake-up time will be described in detail. The model definition unit 13 refers to the user history log at a predetermined timing every day. For example, the time when the operation of the mobile communication terminal 100 was first performed after going to bed in HA or temporary HA or the time when the user moved out of the HA (hereinafter referred to as “HA”) , “The time when the second predetermined trigger occurs”) is used to estimate the latest wake-up time. The model definition unit 13 estimates the wake-up time by dividing the logic into three types as shown in FIG. 9 according to the time zone when the second predetermined trigger occurs.

  Logic3: Wake-up time estimation logic when a second predetermined trigger occurs in a time zone before (μw−σ). Since there are few history logs of the estimated wake-up time in this time zone, in order to increase the certainty of the estimation of the bedtime, the user is inquired of the latest wake-up time at the timing when the wake-up situation occurs, for example. If there is an answer, the time of the answer is estimated as the wake-up time. When there is no answer, or when no wake-up situation has occurred, the wake-up time set by the user is estimated as the wake-up time.

  Logic 4: Wake-up time estimation logic when a second predetermined trigger occurs in the wake-up time zone (μw ± σ time zone) including the average value of the user wake-up time. However, when the number of past estimated wake-up times is 10 or less, 1 hour before and after the wake-up time set in the user set time information is set as the wake-up time zone. When the second predetermined trigger occurs in this wake-up time zone, the time is estimated as the wake-up time.

Logic 5: Wake-up time estimation logic when a second predetermined trigger occurs in a time zone after (μw + σ). On weekdays, the time when the second predetermined opportunity occurs is set as the wake-up time. On holidays, when the second predetermined opportunity occurs, the user is inquired about the wake-up time.
The wake-up time / sleeping time estimated as described above is stored in the user history log, and is used for calculating the average value, estimating the condition of the user, and the like.

(Location information acquisition unit)
The location information acquisition unit 14 acquires location information of the mobile communication terminal 100 carried by the user as the location information of the user. In the present embodiment, the position information acquisition unit 14 is based on the GPS signal received from the position positioning device 300 by the GPS receiver 109 mounted on the mobile communication terminal 100 at predetermined intervals. Calculate location information. The predetermined period for receiving the GPS signal varies depending on the area where the user is staying. For example, when the user stays at the HA, the cycle is 30 minutes, and when the user stays at the WA, the cycle is 15 minutes. In addition, the acquisition method of position information is not limited to GPS, for example, the position information of the base station where mobile communication terminal 100 is located may be regarded as the position information of the user.

(Time information acquisition unit)
The time information acquisition unit 15 acquires time information representing the current time from the internal clock 108 provided in the mobile communication terminal 100. Note that the time information acquisition method is not limited to this, and for example, the time information can be acquired from an external server device.
(Question)
When a predetermined question timing occurs, the question unit 16 receives an instruction from the area definition unit 11 or the like and notifies the user of the question. Here, as a method of notifying the user of the question, image display on the display device 104 or audio output from the voice output device 106 is used. In addition, as types of questions, there are a question for estimating a user's tension (a feeling of elevation), a question for defining an area, a question for estimating a situation, a question for entertaining a user, and the like.

(Situation estimation unit)
The situation estimation unit 17 includes a situation transition model defined by the model definition unit 13, position information acquired by the position information acquisition unit 14, time information acquired by the time information acquisition unit 15, and before the user transition. Estimate the user's current situation based on the situation.
Specifically, the situation estimation unit 17 determines which classification item (wake-up time zone, start-up time) of the user's life cycle is the current time represented by the time information acquired by the time information acquisition unit 15 based on the life cycle information. Time, working time, etc.). The situation estimation unit 17 determines an area corresponding to the position information acquired by the position information acquisition unit 14. In the present embodiment, for example, a radius within 200 m from the center coordinates indicated by the position information is determined to be within the area. When a plurality of areas corresponding to the position information are determined, for example, the priority order is set in advance in the order of HA>WA> favorite area> temporary HA> temporary WA> specific area. Identify one high area.

  Further, the situation estimation unit 17 acquires the latest situation, that is, the situation before the transition, from the user history log. Then, the situation estimation unit 17 acquires the current situation corresponding to the determined life cycle classification item, the determined area, and the acquired situation before the transition from the situation transition model, and acquires the acquired current situation. Is the current situation of the user. The estimated situation is stored in the user history log and is used for the next situation estimation, condition estimation, and the like.

(Estimated by weekdays and holidays)
In addition, the situation estimation unit 17 estimates a user's current situation by using two kinds of situation transition models depending on whether the user falls on a weekday or a holiday today. For example, when the application program is started, default holiday data as shown in FIG. 10 is stored in the scratch pad 123b. The situation estimation unit 17 determines whether today corresponds to a weekday or a holiday for the user based on the weekday holiday data. The default holiday data can be changed by user input. In addition to using the weekday data, it may be determined for the user whether today is a weekday or a holiday based on the “weekday / holiday classification” stored in the latest user history log. Is possible.

(Response classification part)
The response classification unit 18 includes a positive element response that is a response when the user's mood is raised and a negative element response that is a response when the user is depressed. It is determined to be classified. The determination result is used to estimate the user's tension.
Examples of positive element responses include happy, fun, refreshing, refreshing, likes, and delicious.Examples of negative element responses include, for example, sad, painful, lonely, annoying, angry, disliked, bad, painful, etc. Is mentioned.

(Condition index calculation part)
Based on the user history log stored in the user history log storage unit 12, the condition index calculation unit 19 calculates a condition index serving as an index for estimating a condition representing the state of mind and body of the user. The condition index includes, for example, “sleep index”, “labor index”, “energy index”, “mannerin index”, “withdrawal index”, “tension index”, and the like.

(Sleep index)
The “sleep index” is an index that serves as an index as to whether or not the user has enough sleep. The condition index calculation unit 19 is based on the estimated wake-up time and the estimated bedtime stored in the user history log at the timing of the first GPS positioning after midnight, for example, the user's past sleep time (that is, the estimated sleep time). The average value of the time from the bedtime to the estimated wake-up time is calculated. The period for calculating the average value is, for example, the latest one month. And the condition index calculation part 19 updates a "sleep index" to the value which shows that a user's sleep is more sufficient, when a user's latest sleep time is longer than the average value of the past sleep time. On the other hand, when the latest sleep time of the user is shorter than the average value of the past sleep time, the “sleep index” is updated to a value indicating that the user's sleep is insufficient.
In this embodiment, the value of the “sleep index” is three stages of −1, 0, and 1, and the condition index calculator 19 adds 1 to the “sleep index” if the sleep time is longer than the average value, If the sleep time is shorter than the average value, 1 is subtracted from the “sleep index”.

(Labor index)
The “labor index” is an index serving as an index of whether or not the user is working too much. The condition index calculation unit 19 corresponds to situations such as “work in progress”, “overtime”, “work start”, “work end” in the user history log at the timing of the first GPS positioning after midnight, for example. Based on the attached date and time, the average value of the user's past working hours is calculated. The period for calculating the average value is, for example, the latest one month. It should be noted that the average value may be calculated using only the user history log whose weekday / holiday classification is only weekdays. Then, the condition index calculation unit 19 updates the “labor index” to a value indicating that the user's working time is longer when the user's latest working time is longer than the average value of past working hours. On the other hand, if the user's latest working time is shorter than the average value of past working hours, the “labor index” is updated to a value indicating that the user's working time is shorter.
In this embodiment, the value of the “labor index” is three stages of −1, 0, and 1, and the condition index calculation unit 19 subtracts 1 from the “labor index” if the working time is longer than the average value, If the working time is shorter than the average value, 1 is added to the “labor index”.

(Energy Index)
“Energy index” is an index indicating the degree of the user's energy. The condition index calculation unit 19 calculates the “energy index” based on the “sleep index” and the “labor index”.
In this embodiment, the “energy index” is obtained by adding the “sleep index” and the “labor index”, and the value of the “energy index” is five levels from −2 to 2.
(Mannery index)
The “Mannery index” is an index serving as an index as to whether or not there is a change in the daily life pattern of the user.

  The condition index calculation unit 19 is based on the date and time associated with the situations such as “work end” and “return home” in the user history log. Time). The period for calculating the time of returning home is, for example, the last five weekdays. The condition index calculation unit 19 determines that the difference in the calculated time during each return home is all (or may be a part) within a first predetermined time (for example, within 30 minutes). Is updated to a value indicating that the change in the life pattern of the user is smaller. On the other hand, if the difference in time during each day's return to home exceeds one (or all) time difference over a first predetermined time, the “Manneri index” indicates that the change in the user's life pattern is larger. Update to the indicated value.

In addition, the condition index calculation unit 19 uses the user's daily attendance time (that is, from the start of attendance) based on the date and time associated with the situations such as “commencement start” and “work start” in the user history log. Time to start work). The period for calculating the working hours is, for example, the last five weekdays. The condition index calculation unit 19 determines that the difference between the calculated daily working hours is all (or may be a part) within a second predetermined time (for example, within 30 minutes), for example, “Manneri index” Is updated to a value indicating that the change in the life pattern of the user is smaller. On the other hand, if the difference in time during work for each day exceeds one (or all) of the second predetermined time, the change in the life pattern of the user is greater than the “Manneri index”. Update to the indicated value.
In this embodiment, the value of the “manneriness index” is two stages of −1 and 0, the value indicating that the change in the user's life pattern is larger is 0, and the change in the user's life pattern is smaller. The value indicating is -1.

(Withdrawal index)
The “withdrawal index” is an index indicating a tendency of withdrawal in which the user avoids participation in society. Based on the user history log and the area definition information, the condition index calculation unit 19 indicates that the user's situation on holiday is “at home” and the “energy index” is relatively high in the user's energy. And the user stays in the area where the user lives (HA or temporary HA in the present embodiment) for a third predetermined time or more (for example, 6 hours or more). If it is determined that the user is, the “withdrawal index” is updated to a value indicating that the user has a higher tendency to withdraw.

In addition, the condition index calculation unit 19 determines that the user has moved out of the HA and the temporary HA based on the user history log when the “withdrawal index” indicates that the tendency of the withdrawal of the user is higher. If the user's situation is “going out”, the “withdrawal index” is updated to a value indicating that the tendency of the withdrawal of the user is lower.
In this embodiment, the value of the “withdrawal index” is two stages of −1 and 0, and the value indicating that the tendency of the withdrawal of the user is higher is −1, and the tendency of the withdrawal of the user is The value indicating lower is 0.

(Tension index)
The “tension index” is an index indicating whether the user's mood is uplifting or sinking. The condition index calculation unit 19 updates the “tension index” based on the determination result by the response classification unit 18.
In the present embodiment, the value of the “tension index” has three levels of −1, 0, and 1. The condition index calculating unit 19 adds 1 to the “tension index” when the response classifying unit 18 determines that the user response is a positive factor response, and the response classifying unit 18 determines that the user response is a negative factor response. If it is, 1 is subtracted from the “tension index”.

(Serif decision section)
The serif determining unit 20 determines a serif to be presented to the user based on the user's condition index estimated by the condition index calculating unit 19. As a method for determining a serif, for example, the serif determining unit 20 determines a serif corresponding to a condition index by referring to a database in which an appropriate correspondence relationship between a condition and a serif is defined. For example, when the “sleep index” is −1, the line determination unit 20 determines the line “I have recently lacked sleep”.

(Condition estimation program)
As described above, the program stored in the ROM 121 and the application program stored in the application storage area 123a (corresponding to “condition estimation program”) are stored in the mobile communication terminal (corresponding to “computer”) 100 to the user. A user history log storage unit 12 that stores a user history log that is a history of past actions, and a condition index calculation unit 19 that calculates a condition index serving as an index for estimating a user's condition based on the user history log. It is a program to make it function.

(Operation)
Next, a basic flow of the condition estimation process according to the embodiment of the present invention will be described with reference to the flowchart shown in FIG.
First, the user operates the mobile communication terminal 100 to give an instruction to download an application program from the server 200. As a result, the mobile communication terminal 100 downloads the application program and accompanying data from the server 200, stores the application program in the application storage area 123a, and stores the data in the scratch pad 123b.

  Next, the user performs an operation of starting the application program. Accordingly, the mobile communication terminal 100 first displays the initial setting screen on the display device 104 by starting the application program and executing processing according to the program. The user inputs initial setting items interactively from the initial setting screen. Initial setting items include user name, birthday, wake-up time, bedtime, start time, work time, etc., user set time information, weekday classification by day of the week, HA, WA, favorite area, etc. There is an area name.

  The mobile communication terminal 100 performs initial setting based on the input data (step S1). Specifically, the mobile communication terminal 100 stores the input data in the scratch pad 123b. The area definition unit 11 associates the input area name such as HA and WA with the position information of the mobile communication terminal 100 when the area name is input, and stores it in the scratch pad 123b as area definition information. Moreover, the model definition part 13 defines life cycle information based on the input user setting time information. And the model definition part 13 defines the transition order of the user's situation depending on the defined life cycle information and the area, and stores it in the scratch pad 123b as the situation transition model. The situation estimation unit 17 initializes the situation. In the present embodiment, when one or more HA and WA are input by the user, the situation is initially set to “being lost”. After the initial setting, the situation estimation unit 17 is based on the situation transition model, the position information acquired by the position information acquisition unit 14, the time information acquired by the time information acquisition unit 15, and the situation before the user transition. Thus, the user's current situation is sequentially estimated.

  When the situation estimation unit 17 estimates the situation, the user history log storage unit 12 outputs a line to the display device 104 or the sound output device 106 when the model definition unit 13 calculates the latest estimated wake-up time or estimated bedtime. When the user answers the question, when the GPS positioning is performed, when the mobile communication terminal 100 is operated, etc., the date and time are stored in the scratch pad 123b as the user history log ( Step S2).

The condition index calculation unit 19 is based on the user history log stored in the user history log storage unit 12, and serves as an index for estimating the user's condition “sleep index”, “labor index”, “energy index” , Condition indices such as “manneriness index”, “withdrawal index”, and “tension index” are calculated (step S3).
Next, the serif determination unit 20 determines a serif that matches the calculated condition index, and outputs an image on which the determined serif is displayed to the display device 104 or outputs sound from the sound output device 106. (Step S4).

As described above, since the mobile communication terminal 100 calculates the condition index that serves as an index for estimating the user's condition based on the user history log, the user's condition can be easily obtained without using a sensor or a tag. It is possible to estimate.
In addition, since the user history log stores the user situation, wake-up time, bedtime, and the like estimated with high accuracy, the mobile communication terminal 100 calculates the condition index with high accuracy based on the highly accurate user history log. can do.

In addition, since the mobile communication terminal 100 presents the user with an appropriate line corresponding to the user's condition index, it is possible to provide the user with an impression that there is a person watching the user and to provide healing. It becomes.
In the above-described embodiment, the mobile communication terminal 100 has been described as having each function illustrated in FIG. 3, but the present invention is not limited to this, and the server 200 has each function illustrated in FIG. 3. Alternatively, a plurality of devices may have the functions shown in FIG.

  Further, in the above-described embodiment, it has been described that the sleep index, the labor index, the energy index, the manner index, the withdrawal index, and the tension index exist as types of the condition index of the user, but the condition index is limited to this. For example, a “neat index” may exist. For example, the “neat index” may be updated to a value indicating a higher tendency to take a job when the number of days off is a predetermined number of days or more and the user's moving distance is a certain distance or less. Good.

  Further, the method of calculating various condition indexes by the condition index calculating unit 19 described in the above-described embodiment is merely an example. For example, as a method of calculating the “mannerial index”, the occurrence time or end time for each situation for each day is determined, and the difference in the occurrence time or end time for each situation for each determined day is all (or part). If it is within a predetermined time, the “manneriness index” is updated to a value indicating that the change in the user's life pattern is smaller, and even if there is one (or all) difference in the occurrence time or end time for each situation per day It is also conceivable that the “manneriness index” is updated to a value indicating that the change in the user's life pattern is larger when the time is exceeded.

  INDUSTRIAL APPLICABILITY The present invention can estimate the user's condition simply by carrying the user terminal, and can provide various services suitable for the user's condition.

It is a figure which shows the whole structure of the agent system which concerns on embodiment of this invention. It is a block diagram which shows the hardware constitutions of the mobile communication terminal which concerns on the same embodiment. It is a block diagram which shows the function structure of the mobile communication terminal which concerns on the same embodiment. It is a figure showing an example of data composition of a user history log concerning the embodiment. It is a figure which shows an example of the situation transition model which concerns on the embodiment. It is a figure which shows an example at the time of expressing the situation transition model shown in FIG. 5 in a table format. It is a figure which shows an example of the transition pattern of the weekday situation which concerns on the embodiment. It is a figure which shows an example of the transition pattern of the holiday situation which concerns on the embodiment. It is an example of the graph used for estimation of bedtime and wake-up time concerning the embodiment. It is a figure which shows an example of the default weekday holiday data concerning the embodiment. It is a flowchart which shows the flow of the condition estimation process which concerns on the same embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Mobile communication terminal 11 Area definition part 12 User history log storage part 13 Model definition part 14 Position information acquisition part 15 Time information acquisition part 16 Question part 17 Situation estimation part 18 Response classification part 19 Condition index calculation part 20 Serif determination part 102 Storage Device 103 Communication interface 104 Display device 105 Input device 106 Sound output device 107 Input device 108 Internal clock 109 GPS receiver 123 Non-volatile memory 123a Application storage region 123b Scratch pad 200 Server 300 Positioning device

Claims (12)

In the condition estimation device that estimates the condition representing the state of mind and body of the user,
User history log storage means for storing a user history log that is a history of past actions of the user;
A condition estimation apparatus comprising: a condition index calculation unit that calculates a condition index serving as an index for estimating the user condition based on a user history log stored in the user history log storage unit. Stores a situation transition model in which the order of transition of the situation representing the state of the user's behavior depending on the life cycle information representing the daily life cycle of the user and the position information representing the place where the user is located is defined for each holiday. Model definition means to
Position information acquisition means for acquiring position information representing the current location of the user;
Time information acquisition means for acquiring time information representing the current time;
Based on the situation transition model, the position information acquired by the position information acquisition means, the time information acquired by the time information acquisition means, and the situation before the user transition, the current situation of the user is estimated. Further comprising situation estimation means,
The user history log storage unit stores the current situation estimated by the situation estimation unit and the time when the current situation was changed to as the user history log. Condition estimation device. The condition index includes a sleep index that is an indicator of whether or not the user's sleep is sufficient,
The condition index calculating means is:
Based on the user history log, an average value of the user's past sleep time is calculated, and when the latest sleep time of the user is longer than the average value, the sleep index is more sufficient for the user's sleep. It updates to the value which shows that it exists, and when it is shorter than the said average value, the said sleep index | exponent is updated to the value which shows that the said user's sleep is inadequate. Condition estimation device. The condition index includes a labor index that is an index of whether the user is working too much,
The condition index calculating means is:
Based on the user history log, an average value of the user's past working hours is calculated, and when the user's most recent working time is longer than the average value, the labor index is longer than the user's working time. 4. The value representing that is updated, and if it is shorter than the average value, the labor index is updated to a value representing that the working time of the user is shorter. The condition estimation apparatus described in 1. The condition index includes an energy index indicating the degree of energy of the user,
The condition index calculating means is:
The condition estimation device according to claim 4, wherein the energy index is calculated based on the sleep index and the labor index. The condition index includes a tension index indicating the degree of uplift of the user's mood,
The response to the question to the user is classified as either a positive factor response that is a response when the user's mood is uplifted or a negative factor response that is a response when the user's mood is depressed. A response classification means for determining whether or not
The condition index calculating means is:
The condition estimation apparatus according to claim 1, wherein the tension index is updated based on a determination result by the response classification unit. The condition index includes a manner index that is an indicator of whether there is a change in the daily life pattern of the user,
The condition index calculating means is:
Based on the user history log, the time when the user's situation is returning from work to returning home is calculated every day,
If the calculated difference in the time of returning home is within the first predetermined time, the manner index is updated to a value indicating that the change in the user's life pattern is smaller,
When the difference in the time of going home every day exceeds the first predetermined time, the manner index is updated to a value indicating that a change in the life pattern of the user is larger. The condition estimation apparatus according to any one of claims 2 to 6. The condition index calculating means is:
Based on the user history log, calculate the time during which the user's situation is working from the start of work to the start of work for each day,
If the calculated difference in working hours per day is within a second predetermined time, the manner index is updated to a value indicating that the change in the user's life pattern is smaller,
When the difference in the working hours for each day exceeds the second predetermined time, the manner index is updated to a value indicating that a change in the life pattern of the user is larger. The condition estimation apparatus according to claim 7. Further comprising area definition means for storing area definition information representing a correspondence relationship between the location information and an area representing a place where the user stays according to the purpose;
The condition index includes a withdrawal index indicating a tendency of withdrawal in which the user avoids participation in society,
The condition index calculating means is:
Based on the user history log and the area definition information, the situation of the user on a holiday is at home, and the energy index indicates that the degree of energy of the user is relatively low, and the user When it is determined that the user stays in the area where the user resides for a third predetermined time or more, the withdrawal index is updated to a value indicating that the withdrawal tendency is higher. The condition estimation apparatus according to any one of 2 to 8. The condition index calculating means is:
When it is determined that the user has moved out of the area in which he lives based on the user history log, the withdrawal index is updated to a value indicating that the withdrawal tendency of the user is lower. The condition estimation apparatus according to claim 9.   The condition according to any one of claims 1 to 10, further comprising a serif determining unit that determines a serif to be presented to the user based on the condition index calculated by the condition index calculating unit. Estimating device. Computer
User history log storage means for storing a user history log that is a history of past actions of the user;
For functioning as a condition index calculating means for calculating a condition index serving as an index for estimating a condition representing the state of mind and body of the user based on the user history log stored in the user history log storage means Condition estimation program.

Images