KR100788922B1 - Method and system for offering intelligent home service - Google Patents

Abstract

An intelligent home service supply method and a system thereof are provided to increase accuracy in prediction of a home service by comparing results predicted by two methods, and to obtain a user's biological and environmental contexts, thereby predicting the home service from the obtained contexts. User biological context data and environmental context data are obtained from a sensor device(S110). The obtained context data is normalized(S120). The normalized context data is inputted to SVM(Support Vector Machine) algorithm to predict a home service, and the home service is predicted according to rules made by data mining for the normalized context data(S130). If the two predicted results are the same, the home service is carried out according to the results(S140,S150).

Description

Method and system for providing intelligent home service {Method and system for offering intelligent home service}

1 is a flowchart illustrating a method for providing an intelligent home service according to an embodiment of the present invention.

2 is a diagram illustrating obtaining a living body and an environment context from a sensor device according to an embodiment of the present invention.

3 is a diagram illustrating a context normalization table according to an embodiment of the present invention.

4 is a diagram illustrating a structure for predicting a home service using a linear support vector machine (SVM) algorithm according to an embodiment of the present invention.

5A is a diagram illustrating a distribution of TV operation with time according to a user's learning.

5B is a diagram illustrating a TV operation distribution according to a pulse according to a user's learning.

6 is a flowchart illustrating a step of recommending and providing a home service according to an electrocardiogram analysis according to an embodiment of the present invention.

7 is a diagram illustrating a normal ECG signal obtained from an ECG sensor and an ECG signal out of a normal range.

8 is a block diagram of an intelligent home service providing system according to an embodiment of the present invention.

<Explanation of symbols for the main parts of the drawings>

10: biometric context 15: environmental context

20: SVM algorithm 210: sensor unit

220: context management unit 230: history-based prediction unit

240: rule-based prediction unit 250: prediction result comparison unit

260: home service control unit 270: ECG analysis unit

280: Home Service Recommendation

The present invention relates to providing an intelligent home service, and more particularly, an intelligent home that obtains a living body and environment context from a sensor device and provides a home service to a user through a two-step home service prediction process. It relates to a service providing method and system.

Recently, research on the home network environment is being actively conducted, and in particular, research on the home service that operates intelligently to enable mutual communication with human beings is being actively conducted.

Smart homes that provide automatic and efficient services for humans are realized by adding intelligent smart middleware to homes with ubiquitous computing.

Research on such middleware includes Microsoft's Easy living, Aware Home by AHRI (Aware Home Research Initiative), and Adaptive House.

Easy living suggests tracking a user's location and providing home appliances based on the user's location. However, Easy living mainly focuses on user location detection and user authentication, and does not contain information on intelligent home service provision by analyzing user and environment contexts.

On the other hand, Aware Home focuses on constructing an intelligent home suitable for the elderly who live alone, and provides appropriate services. For example, Aware Home provides a service that generates an alarm when there is no user movement for a certain period of time. In addition, Adaptive House proposed a method for maximizing user satisfaction while minimizing energy consumption. However, this study does not contain information on providing intelligent home services based on user and environment contexts.

That is, in the past, there have been various studies for implementing a smart home, but studies on providing a home service by predicting a user's behavior pattern by performing a situation analysis based on the user's biometric information and environment information are insufficient. The accuracy of was not high.

The present invention has been devised to improve the above problems, and the technical problem to be achieved by the present invention is to provide an intelligent home service that automatically executes a home service through a two-step home service prediction step from the user's biological and environmental contexts. There is.

The objects of the present invention are not limited to the above-mentioned objects, and other objects that are not mentioned will be clearly understood by those skilled in the art from the following description.

In order to achieve the above object, the intelligent home service providing method according to an embodiment of the present invention comprises the steps of (a) obtaining user biometric context data and environment context data from the sensor device, (b) the obtained context Normalizing data, (c) inputting the normalized context data into a predetermined algorithm to predict a home service, predicting the home service according to a predetermined rule, and (c) If the two predicted results in are the same, then executing the home service according to the result.

In order to achieve the above object, the intelligent home service providing system according to an embodiment of the present invention is a sensor unit for obtaining user biometric context data and environment context data from a sensor device, a context for normalizing the obtained context data A management unit, a history-based prediction unit predicting a home service by inputting the normalized context data into a predetermined algorithm, a rule-based prediction unit predicting a home service according to a predetermined rule, and the history-based prediction unit And a prediction result comparison unit comparing the results predicted by the rule-based prediction unit, and a home service control unit controlling a home service according to the predicted result when the result predicted by the prediction result comparison unit is the same.

Specific details of other embodiments are included in the detailed description and the drawings.

Advantages and features of the present invention and methods for achieving them will be apparent with reference to the embodiments described below in detail with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but can be implemented in various different forms, and only the embodiments make the disclosure of the present invention complete, and the general knowledge in the art to which the present invention belongs. It is provided to fully inform the person having the scope of the invention, which is defined only by the scope of the claims. Like reference numerals refer to like elements throughout.

Hereinafter, the present invention will be described with reference to the drawings for explaining an intelligent home service providing method and system according to embodiments of the present invention.

Before the term is defined, the context is defined as information that can characterize the state of the object, where the object can be a person, place, or physical object. For example, it may be body temperature, pulse rate, room temperature, facial expression, human position, posture, and the like.

1 is a flowchart illustrating a method for providing an intelligent home service according to an embodiment of the present invention.

 First, biometric context 10 and environmental context 15 data of a user are obtained from a sensor device such as a temperature sensor, a camera, a pulse rate, and a temperature sensor (S110). Details will be described later with reference to FIG. 2.

Next, the context 10 and 15 data obtained from the sensor device is normalized from the preset context normalization table (S120). That is, the data of each context 10 and 15 may vary in units and ranges, and normalize all data in a specific range of each context 10 and 15 to values such as 0.1, 0.2, and 0.3. Details of the context normalization table will be described later with reference to FIG. 3.

After normalizing context 10 and 15 data, the context 10 and 15 data is input to a predetermined algorithm to predict a home service (S130). At this time, the predetermined algorithm is a context input based on a plurality of data stored about the behavior pattern of the user through learning (for example, context data such as how much the user's pulse is, how the posture is, etc. when the user turns on the TV). This algorithm finds the home service content that best matches (10, 15). Preferably, a support vector machine (SVM) algorithm 20 may be used.

In addition, the home service is predicted based on the normalized context 10 and 15 data according to a predetermined rule (S130). The predetermined rule refers to a rule that activates a specific home service when context 10 and 15 data in a specific value range is input. Preferably, the home service is predicted according to the rules created by the data mining technique.

It is determined whether the results predicted by the two methods described above are the same (S140). Only when the predicted result is the same, the home service is executed according to the predicted result (S150). If the predicted result is not the same, the process returns to the beginning again and receives the context (10, 15) data (S110) and repeats the process of predicting the home service again.

In this case, the home service may be related to the operation of the TV, audio, DVD, air conditioner, light, and the like, and more specifically, the selection of a TV channel, the type of music, the movie genre, the intensity of the air conditioner, the brightness of the light, and the like may be considered. .

2 is a diagram illustrating obtaining (S110) a living body 10 and an environment context 15 from a sensor device according to an embodiment of the present invention.

The contexts 10 and 15 according to an embodiment of the present invention are five biological contexts 10 including a pulse, a body temperature, an electrocardiogram (ECG), a facial expression, a user's posture, and an environmental feature of a room. It consists of three environmental contexts 15: room temperature, user location, and time. It will be appreciated by those skilled in the art that the contexts 10 and 15 can represent not only the eight listed but also the biological and environmental features of the user and can be variously selected if obtained by the sensor device.

The eight contexts listed above may be obtained through the respective sensor devices (S110). The room temperature is from the temperature sensor, the eye focus is from the eye focus camera, the electrocardiogram is from the ECG sensor, the facial expression is from the facial expression detection camera, and the pulse and body temperature are connected to the PDA (Personal Digital Assistant). Through the sensors, the position and posture of a person can be known through a plurality of cameras installed in each area of the room. It will be appreciated that the sensor device for detecting the contexts 10 and 15 may be variously selected in consideration of cost and performance. Data of each context 10 and 15 obtained from the sensor device is then subjected to a normalization process.

3 is a diagram illustrating a context normalization table according to an embodiment of the present invention.

Since the data of each context 10 and 15 varies in units and ranges, as shown in FIG. 3, the context 10 and 15 data is normalized to nine values of 0.1 to 0.9 in a specific range of data (S120). ). In the table, the user's pulse was normalized to a total of nine values, which means that the average adult pulse at rest is about 72 times per minute, and pulses below 40 and pulses above 140 correspond to the actual user's emergency. We exclude because we do. In the case of body temperature, the normal body temperature of a person is normalized to 9 levels in this range, considering that there are differences between ages and individuals, but usually 34 to 39 degrees. For facial expressions, see six basic facial expressions (Surprise, Fear, Sad, Angry) by Charles Darwin (see "The expression of the emotions in man and animals", Electronic Text Center, University of Virginia Library). , Disgust, Happy) and no expression (Blank). In the case of room temperature, the normalized temperature is normalized to 9 steps based on the fact that the average room temperature is 19 to 22 degrees. The time was normalized into nine steps as shown in the table, taking into account the time spent by humans. The user location was normalized by dividing the room into nine zones. The normalization table of FIG. 3 is just an example according to an embodiment of the present invention, and the normalization process may not only be divided into nine values, but also may further divide the data in more detail to normalize to more values or to normalize to less values. to be.

When normalizing the context (10, 15) data, the context (10, 15) data obtained from the sensor device is received at regular intervals. If the input data does not have a value corresponding to the range of values defined in the above-described context normalization table, error processing is performed and new context (10, 15) data is input again. The context (10, 15) data that has passed the error check is normalized (S120) according to a preset context normalization table, and the normalized data is transmitted to the step of predicting a home service (S130).

The normalized data is input to a predetermined algorithm to predict the home service (S130). Preferably, a support vector machine (SVM) algorithm 20 is used. FIG. 4 is a diagram illustrating a structure for predicting a home service using the SVM algorithm 20 according to an embodiment of the present invention, according to a location of a user using a plurality of linear support vector machines (LSVMs). Run hierarchically to predict home services. By inputting the normalized context (10, 15) to the SVM algorithm 20, the user's behavior pattern that best matches the input context (10, 15) data based on the data on the user's behavior pattern by the conventional learning Predict.

In addition, the normalized context 10 and 15 data is predicted to the home service according to a predetermined rule (S130), preferably the home service is predicted according to a rule created by a data mining technique. .

Data mining is the process of discovering correlation rules in data from a database, but not in the past, finding new data models, extracting actionable information in the future, and using it in decision making.

FIG. 5A is a diagram illustrating a distribution of TV operation according to time according to the learning of the user, and FIG. 5B is a diagram of a distribution of TV operation according to the pulse according to the learning of the user.

Referring to FIG. 5A, it can be seen that many times a user operates a TV between 17 and 21 o'clock. In other words, the relationship between time and the operation of the TV can be seen that the user is likely to be satisfied if you operate the TV between 17 and 21 o'clock. However, looking at the graph, it can be seen that there is a lot of data that does not operate the TV between 17 o'clock and 21 o'clock. Thus, another context 10, 15 that affects the operation of the TV is further considered. Looking at Figure 5b, it can be seen that the operation of the TV often appears when the user's pulse is between 82 ~ 100. In this way, the TV can be activated automatically when the user's pulse is 82-100 between 17 and 21 o'clock. This rule may not be enough to accurately predict a user's behavior pattern, but if you correlate more contexts (10, 15) to create a rule, you can create a more accurate rule.

For example, if the pulse is between 80 and 100, the body temperature is lower than 37.0 degrees, and the facial expression is sad, play soft music. Can produce

Only when the results predicted by the SVM algorithm 20 and the data mining rule are the same, the home service is executed according to the predicted results (S150). Therefore, it is possible to more accurately predict the behavior of the user to provide a home service.

TABLE 1

TV DVD audio Estimated Execution Count with SVM Algorithm 123 156 211 Execution instruction count predicted by rule by data mining 148 111 90 The same number of instructions from both methods 76 90 107 The number of commands rejected by the user 21 24 31 Execution command accuracy 72.3% 73.3% 71.0%

Table 1 shows the accuracy of the predicted results shown according to one embodiment of the invention. As a result of providing the home service when the result predicted by the two methods is the same, it can be seen that the accuracy of the prediction result is more than 70%. It is expected that the more accurate the content of the context 10, 15 will be, the more diverse it will be.

6 is a flowchart illustrating a step of recommending and providing a home service according to an electrocardiogram analysis according to an embodiment of the present invention.

After the home service is executed according to the result predicted by the above method, the stress change of the user is analyzed through electrocardiogram (ECG) analysis (S160). ECG is a curve recording the active current according to the contraction of the heart, it can be determined that the amount of stress is large and small according to the narrow and wide PR interval of the ECG. 7 is a diagram illustrating a normal ECG signal obtained from an ECG sensor and an ECG signal out of a normal range. Can be.

When it is determined that the stress of the user is alleviated, the context (10, 15) data for predicting the corresponding home service may be stored and used as important data for predicting the home service from the other contexts (10, 15). In addition, if stress is alleviated, a home service having a similar tendency to the provided home service may be recommended and provided to the user (S170).

8 is a block diagram of an intelligent home service providing system according to an embodiment of the present invention.

The intelligent home service providing system according to an embodiment of the present invention includes a sensor unit 210, a context manager 220, a history-based predictor 230, a rule-based predictor 240, a prediction result comparer 250, and It may include a home service control unit 260. The ECG analyzer 270 and the home service recommender 280 may further be included.

The sensor unit 210 obtains the user's biological context 10 and environment context 15 data from a sensor device such as a temperature sensor, a camera, a pulse and a temperature sensor. In this case, the biological context 10 data may be a pulse, body temperature, electrocardiogram (ECG), facial expression, eye focus, user posture, and the like. The environment context 15 data may include user location, room temperature, time, and the like.

The context manager 220 normalizes the context 10 and 15 data obtained from the sensor device from a preset context normalization table.

The history-based predictor 230 inputs the normalized context 10 and 15 data input from the context manager 220 to a predetermined algorithm to predict a home service for a user. In this case, the predetermined algorithm is an algorithm for finding home service contents that best match the input contexts 10 and 15 based on a plurality of stored data regarding a user's behavior pattern through learning. Certain algorithms preferably use SVM (Support Vector Machine) algorithm 20.

The rule-based predictor 240 inputs normalized context 10 and 15 data input from the context manager 220 to a predetermined rule to predict a home service for a user. do. The predetermined rule refers to a rule that activates a specific home service when context 10 and 15 data in a specific value range is input. Preferably, the home service may be predicted according to a rule created by a data mining technique.

The prediction result comparison unit 250 compares the results predicted by the above two methods. If the predicted result is the same, the predicted result is transmitted to the home service controller 260. If the predicted result is not the same, the context 10 and 15 data are input from the sensor unit 210 and the above-described process is repeated.

The home service controller 260 executes a home service according to the predicted result when the predicted result comparison unit 250 determines that the predicted results by the two methods are the same.

Stress after the home service is executed in the ECG analyzer 270 and the ECG analyzer 270 analyzing the stress change of the user by analyzing the ECG before and after the home service is executed by the home service controller 260. If it is determined that the mitigation may further include a home service recommendation unit 280 for recommending or providing other home services of a similar tendency.

In this case, the term '~ part' means a hardware component such as software, a field programmable gate array (FPGA), or an application specific integrated circuit (ASIC), and performs certain functions. However, it is not meant to be limited to software or hardware. The component may be configured to be in an addressable storage medium and may be configured to play one or more processors. Thus, as an example, the component may include components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, procedures, Subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functionality provided by the components can be combined into a smaller number of components or further separated into additional components. In addition, the components may be implemented to regenerate one or more CPUs in a device.

Those skilled in the art will appreciate that the present invention can be embodied in other specific forms without changing the technical spirit or essential features of the present invention. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive. The scope of the present invention is indicated by the scope of the following claims rather than the above description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents are included in the scope of the present invention. Should be interpreted.

According to the intelligent home service providing system and method of the present invention as described above has one or more of the following effects.

First, the accuracy of prediction of home service can be improved by comparing the results predicted by the two methods.

Second, there is an advantage that the home service can be predicted from the biological and environmental context of the user.

Claims (12)

(a) obtaining user biological context data and environmental context data from the sensor device; (b) normalizing the obtained context data; (c) predicting a home service by inputting the normalized context data into a support vector machine (SVM) algorithm and predicting the home service according to a rule made by data mining; And (d) if the two predicted results in step (c) are the same, executing the home service according to the result. delete delete The method of claim 1, The biometric context data includes information about pulse rate, body temperature, electrocardiogram (ECG), facial expression, eye focus and user posture. The method of claim 1, And the environment context data includes information about user location, room temperature and time. The method of claim 5, (e1) analyzing the stress change of the user by the ECG analysis before and after the home service is executed by the step (d); And (e2) recommending or providing another home service having a similar tendency when it is determined that the stress is alleviated after the home service is executed in the step (e1).  A sensor unit configured to obtain user biological context data and environment context data from the sensor device; A context manager which normalizes the obtained context data; A history-based prediction unit for predicting a home service by inputting the normalized context data into a support vector machine (SVM) algorithm; A rule-based prediction unit for predicting a home service according to a rule generated by data mining of the normalized context data; A prediction result comparison unit comparing the results predicted by the history-based prediction unit and the rule-based prediction unit; And And a home service control unit controlling a home service according to the predicted result when the result predicted by the prediction result comparison unit is the same. delete delete The method of claim 7, wherein The biometric context data includes information about pulse rate, body temperature, electrocardiogram (ECG), facial expression, eye focus and user posture. The method of claim 7, wherein The environment context data includes information about user location, room temperature and time. The method of claim 10, An electrocardiogram analyzer for analyzing a stress change of a user by analyzing the electrocardiogram (ECG) before and after the home service is executed by the home service controller; And And a home service recommendation unit for recommending or providing another home service having a similar tendency when the ECG analyzer determines that the stress is alleviated after the home service is executed.

Images