KR101319603B1 - Adaptive Context-aware System for Recommendation Service Improvement on Mobile Cloud - Google Patents

Abstract

Active situation recognition system for recommendation service in mobile cloud environment that collects situation information from sensors in mobile cloud environment, constructs user profile and situation information with ontology, and searches and provides services corresponding to sensed situation information. A sensor processor comprising: a sensor processor configured to receive raw data collected from sensors; A situation information collector for generating situation information by converting the raw data into a data format defined by an ontology; A situation engine for performing redundancy check on the situation information and performing inference using the situation information and user profile information; A knowledge base manager for managing the situation information and the user's profile information ontology; And a service analysis module for searching for context information and ontology of the user derived from the context engine to find a service for the user.
By the active situation recognition system as described above, a suitable service can be provided by recommending a more suitable and efficient service for each user's situation in a mobile cloud environment.

Description

Active Context-aware System for Recommendation Service Improvement on Mobile Cloud}

The present invention collects context information from a sensor in a mobile cloud environment, constructs a user profile and context information ontology, searches for and provides a service corresponding to the sensed context information, and is active for a recommendation service in a mobile cloud environment. A situational awareness system.

The increase in the variety and performance of mobile devices has led to a surge in the number of mobile device users. Improvements in the mobile network infrastructure have allowed for increased data throughput and increased data rates. The development of the mobile devices and the improvement of the mobile network infrastructure have made the exchange and sharing of information and resources using the mobile devices more active, and the use of web services through the mobile devices is also increasing.

In line with this trend, many researches are being actively conducted on the "Mobile Cloud" service which combines cloud computing technology with existing mobile services. Mobile cloud computing allows users to access and use resources and services beyond their own mobile cloud without leaving their network on the go with their mobile devices.

  In such a mobile cloud computing environment, the importance of a personalization service that can provide appropriate resources and services to the user by using information such as the current situation, behavior, propensity, and preference of the user is becoming more important. In order to provide personalized services, systems that provide services can analyze the user profile such as user needs, behaviors, and preferences along with the awareness of the situation in which the user is in a position to recommend services that meet the request. It should be possible.

However, the existing Recommendation Services for personalization services are mostly web-based desktop services and are not suitable for mobile devices. In addition, since there are many services that consider only the location information among the user's context information, consideration of other context information other than the user's location information is very insignificant. We do not recommend and provide suitable services.

Therefore, in consideration of the situational awareness of users in the mobile cloud environment and the preferences of users, it searches for various and widely distributed services and information on the network to recommend more efficient and context-sensitive personalized services for each user. There is also a need for a framework that can provide services and support services that provide web resources through the mobile cloud.

An object of the present invention is to solve the problems described above, to collect the situation information from the sensor in the mobile cloud environment, to construct the user profile and the situation information as an ontology, search for the service corresponding to the sensed situation information It is to provide an active situation recognition system for the recommendation service in the mobile cloud environment.

In order to achieve the above object, the present invention relates to an active situation recognition system for a recommendation service in a mobile cloud environment, comprising: a sensor processor for receiving raw data collected from sensors; A situation information collector for generating situation information by converting the raw data into a data format defined by an ontology; A situation engine for performing redundancy check on the situation information and performing inference using the situation information and user profile information; A knowledge base manager for managing the situation information and the user's profile information ontology; And a service analysis module for searching for context information and ontology of the user derived from the context engine to find a service for the user.

In addition, the present invention is an active situation recognition system for a recommendation service in a mobile cloud environment, the situation information collector, a stream processor for converting the raw data to digital; And a sensor data converter for converting the digitally converted data into a data format defined by the ontology.

In addition, the present invention is an active situation recognition system for a recommendation service in a mobile cloud environment, the situation engine, an effective processor for checking the redundancy of the context information; And an inference engine that makes an inference using the situation information and the user's profile information that has passed through the valid processor.

In addition, the present invention is an active situation recognition system for a recommendation service in a mobile cloud environment, the effective processor is characterized in that the redundancy check using the tolerance (hereinafter offset) of the situation information.

In addition, the present invention is characterized in that in the active situation recognition system for the recommended service in the mobile cloud environment, the valid processor discards the situation information when it is determined that the situation information is duplicated.

In addition, the present invention is an active situation recognition system for a recommendation service in a mobile cloud environment, the inference engine is characterized in that it uses a rule-based reasoning system of axiomatic method.

In addition, the present invention is an active situation recognition system for a recommendation service in a mobile cloud environment, the rule used in the inference engine is characterized in that it is defined by a rule markup language (RuleML, Rule Markup Language).

In addition, the present invention is an active situation recognition system for a recommendation service in a mobile cloud environment, the knowledge base manager, a profile handler for managing user profile information; And an ontology handler that manages the situation information based on the ontology.

In addition, the present invention is an active situation recognition system for recommendation services in a mobile cloud environment, the situation information is characterized in that any one or more of the pulse, body temperature, blood pressure.

As described above, according to the active situation recognition system according to the present invention, the effect that the appropriate service can be provided by recommending a more suitable and efficient service for each user's situation in the mobile cloud environment.

According to the active situation recognition system according to the present invention, it is possible to provide a service suitable for the user through the situation awareness and inference of the user by using the context information input by the situation engine, the user's profile information, and the rules defined in the ontology. Effect is obtained.

In particular, in the mobile cloud environment, if personal information such as a personal medical record is required according to the situation, it may be delivered to a mobile device or a linked hospital system and used to provide an appropriate service.

In addition, when comparing the present invention and the prior art Context Toolkit, CoBrA, Gaia, SOCAM, etc., CoBrA, SOCAM, and the present invention use the ontology-based OWL in processing the situation, only SOCAM and the present invention are mobile services. In addition, the framework using the inference engine or the inference engine and knowledge base in service discovery is SOCAM, CoBrA, the present invention, and the frame using the profile considering the user's preference or propensity as historical context information. It can be seen that the workpiece is only the system of the present invention.

 In view of these features, the active situation recognition system according to the present invention can provide a more suitable and accurate recommendation service to a user using a mobile cloud than the existing situation recognition system.

1 is a diagram showing an example of the overall system configuration for implementing the present invention.
2 is a block diagram of a configuration of a system according to an embodiment of the present invention.
3 is a diagram illustrating a process of processing and transferring situation information acquired from a sensor according to the present invention.
4 illustrates a situation information redundancy check algorithm according to the present invention.
5 shows an example of an inference rule using RuleML according to the present invention.
6 shows a flowchart for inferring the user's situation in accordance with the present invention.
7 illustrates a situation inference algorithm according to the present invention.
8 shows an example of a configuration diagram of a user profile according to the present invention.
9 illustrates an example of an ontology for a user profile according to the present invention.
10 illustrates an example of a situational awareness model definition according to the present invention.
11 illustrates an example of a sensor ontology configuration according to the present invention.
12 illustrates an example of a disease data ontology configuration according to the present invention.
Figure 13 shows an example of the situation model configuration diagram according to the present invention.
14 illustrates an example of a situation-specific provision service according to the present invention.
15 illustrates a process for providing a mobile cloud service through situation inference according to the present invention.
Figure 16 shows the reference value of the sensor data for each situation according to the health state according to the experiment of the present invention.
17 illustrates knowledge modeling according to an experiment of the present invention.
18 illustrates a user interface according to an experiment of the present invention.
19 is a comparative table of the present invention and the prior art according to the experiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the drawings.

In the description of the present invention, the same parts are denoted by the same reference numerals, and repetitive description thereof will be omitted.

First, the configuration of the entire system for implementing the active situation recognition system according to the present invention will be described with reference to FIG.

As shown in FIG. 1, the entire system for implementing the present invention is composed of a sensor 10, a mobile 11, a situation recognition system 30, and the like in a cloud environment 20. Also included is a cloud database 40 for storing data in a cloud environment.

The mobile 11 refers to a terminal using a wireless network (or a mobile communication network) such as a smartphone. In particular, the mobile 11 is a terminal capable of wireless internet. The mobile 11 may be connected to the situation recognition system 30 and the cloud network to perform data communication with each other.

The sensor 10 refers to a sensor that is scattered around and senses environmental conditions to collect data. As an example, the sensors 10 are biological sensors that sense and collect biological data such as pulse, body temperature, and blood pressure. The sensor 10 is connected to the situation recognition system 30 and the cloud network and directly transmits the sensed data to the situation recognition system 30 or through the mobile 11. Alternatively, the sensor 10 is installed in the mobile device 11.

The situation recognition system 30 stores the user's profile in advance and collects situation information from the sensor 10. In addition, the situation recognition system 30 finds a service suitable for the user's profile and the surrounding situation and provides the corresponding service to the user or through the mobile 11.

Next, the configuration of an active situation recognition system for a recommendation service in a mobile cloud environment according to an embodiment of the present invention will be described in more detail with reference to FIG. 2.

In the present invention, an active situation for improving a recommendation service technique using a mobile device in a mobile cloud computing environment that provides a more suitable service by using frequently used service information, user personal information, and preference stored profiles and situation information generated to the user. Recognition system. The configuration of an active situation recognition system for improving the recommended service technique is shown in FIG. 2.

As shown in FIG. 2, the active situation recognition system 30 according to an embodiment of the present invention includes a sensor processor 31, a situation information collector 32, a situation engine 33, a knowledge base manager 34, and It consists of a service analysis module 35. In addition, it may be configured to further include a cloud database 40 for storing data.

The sensor processor 31 collects information from sensors scattered around and transfers the information to the situation information collector. The context information collector 32 includes a sensor data converter 32a and a stream processor 32b for digitally converting the received data to be recognized by a computer. The situation engine 33 receives the data from the sensor data converter 32a and uses the valid processor 33a for performing data redundancy check, the situation information received through the valid processor, the profile information of the user, and the situation information modeling. And an inference engine 33b for inferring the situation. In addition, the knowledge base manager 34 is composed of a profile for storing the user's personal information, preferences and situation information, and situation awareness modeling for storing, managing, and providing information necessary for the situation awareness application. . The service analysis module 35 analyzes the required service using the inferred situation.

Specifically, the sensor handler 31 receives raw data collected from the sensors 10 scattered around and transfers the raw data to the stream processor 32a of the situation information collector 32.

Various contextual information generated from an individual user may be obtained through various sensors 10 and profiles of the user.

The contextual information is preferably limited to specific types of data such as biometric data corresponding to personalization service among various sensor information. The reason for the need for such situation information is that situation information varies depending on the situation. That is, if the field used is different, the meaning of the situation information may be different. Therefore, it is necessary to restrict the situation information according to the application.

As an example, the situation information is set to three types of biometric data: pulse, body temperature, and blood pressure. Pulse, body temperature, and blood pressure, which are used as contextual information, are information that is frequently used as an indicator of personal health in the medical field. Contextual normals for pulse, body temperature, and blood pressure are defined using situational rules. Although body temperature is based on rectal temperature, it is not practical to attach a sensor to the rectum for sensing, so it replaces biometric data with skin temperature instead of rectal temperature. Through this normal value, it is possible to determine the validity of such data by collecting situation information outside the normal value range.

Next, the context aggregator 32 converts the raw data into a data format defined by the ontology to generate context information.

As shown in FIG. 3, the context information collector 32 is a stream processor 32a for digitally converting data (or raw data) transmitted from the sensor processor 31 and data that is digitally converted in the ontology. It consists of a sensor data converter 32b for converting to a format.

3 shows the data obtained from the sensors 10 is transferred to the stream processor 32a by the sensor processor 31 and back to the sensor data converter 32b.

The stream processor 32a is composed of a plurality of sensor stream processors that are responsible for converting the sensed situation information from the sensor processor 31 and converting the digital information to the sensor data converter 32b.

The sensor data converter 32b converts the digitally converted data received from the stream processor 32a into the data format defined in the ontology before transmitting the data to the valid processor 33a of the situation engine 33. In charge of.

Next, the context engine 33 performs redundancy check on the context information and performs inference using the context information and the profile information of the user.

The context engine 33 uses the valid processor 33a for redundancy check on the context information received from the sensor data converter 32b, the context information from the valid processor, and the user's profile information. It consists of the inference engine 33b.

The valid processor 33a is responsible for performing redundancy check on the situation information transmitted from the sensor data converter 32b.

Since sensor data is transmitted continuously, if you want to store all of them, you have to spend a lot of storage space and consume processing resources for storage. This is a big problem in the situation information processing. Therefore, in selecting the situation information to be stored, the validity processor 33a executes the following redundancy check to ignore the duplicated data.

Duplicate data can be regarded as the same data value and not much different from previous data. The process of retrieving the same data is not difficult, but it does not differ so much that it is subtle to determine if it is a duplicate or not.

The difficulty of this decision is to use offsets for pulse, body temperature and blood pressure. This offset means a tolerance, and the redundancy check algorithm is shown in FIG. If it is determined that the situation information is duplicated through the redundancy check algorithm, the situation information is discarded. Although the redundancy check is very simple, it is effective in managing the situation information because redundancy is determined for the data within the error. Data that passes the redundancy check is selected as the final status information. The data is transmitted to the ontology processor and stored in the data ontology.

The inference engine 33b utilizes the data received from the validity processor 33a and the context-aware modeling and profile in the knowledge base manager 34. It is responsible for inferring the user's situation.

The reasoning engine 33b uses Bosssam, an axiomatic reasoning system with ontology rule reasoning. The Bossam inference engine, developed for the first time in Korea, was developed to support all of the W3C's semantic web standards, and features software developers to easily define ontology and rules and develop various application services. It also supports time data representation and processing, and allows arithmetic expressions to be written in terms of predicates. Bossam's reasoning engine is a rule-based reasoning system that has the advantages of small program size, light weight, and fast reasoning speed.

In one embodiment of the present invention, a rule markup language (RuleML, Rule Markup Language) was used as a language for defining a rule [Non-Patent Document 1], and the defined rule is as shown in FIG. In the example of FIG. 5, the situation is divided into 'daily or athletic' and 'emergency'.
[Non-Patent Document 1] http://www.ruleml.org/

FIG. 6 is a diagram illustrating a process of deriving a user's situation using the rule and profile information and sensor information defined in FIG. 5.

As shown in FIG. 6, the user's situation is inferred according to the frequency of exercise at the place where the exercise is stored in the user's profile, the probability of exercising at the place, and whether the sensor value collected at the place is within the daily range. do. An algorithm for inferring such a user situation is as shown in FIG. 7.

Next, the knowledge base manager 34 manages the situation information and the user's profile information by ontology.

The knowledge base manager 34 is composed of a profile handler 34a that manages user profile information, and an ontology handler 34b that manages context information on an ontology basis. In order to provide a personalized service suitable for a user in a mobile cloud environment, a storage for storing and using various situation information and personal information generated for a user is required.

The profile handler 34a is responsible for storing and managing the health information, medical treatment information, and other personal information of the individual user and the sensed information, and also defines a profile as a personal information repository.

A profile is written in a language that can be understood by a machine, and most profile information can be prepared in advance, and a service can be adapted to a user's environment. Among them, the user preference profile describes information related to the user and includes the static and dynamic information of the user. When providing a personalization service using a user profile, the user's preference information may be processed to provide information suitable for the user.

A user profile depicts information related to a user and includes static information of a user (such as user terminal capability, user preferences, basic user information, etc.) and dynamic information (current location, surrounding environment, etc.).

The profile handler 34a uses the ontology to construct a personal profile as shown in FIG. 8. 9 is an OWL schema of a user profile for storing user information.

The ontology handler 34b models and configures contextual information based on the ontology.

The construction of ontology-based contextual information model uses the concept of class and instance to structure information obtained from sensors or computing resources into ontology-based context. An instance created through a class inherits all the attributes defined in the class. All properties of the class defined in the ontology are inherited by its instance. At this point, the instance will actually have data from sensors or computing resources. The class also contains an attribute, which contains the actual sensor data. Ontology for contextual information modeling for a user is defined as shown in FIG.

Ontology-based modeling of each contextual information is as follows.

First, the sensor ontology models the basic information of the sensor. 11A is a configuration diagram of a sensor ontology. Sensor ontology consists of one class and four attributes. Each meaning is the same as in FIG. 11B. 11C is an OWL (Ontology Web Language) schema of a sensor ontology for storing sensor information.

Next, in order to filter the sensing information, the ontology of the disease data as shown in FIG. 12A is configured, and the ontology as shown in FIG. 12B is constructed. It is used to filter time intervals and priorities in the situation filtering module using disease data ontology.

Next, the situation information of the user is composed of a situation, a device, a profile, a disease, a service, an exercise, a sensor, and the like, and FIG. 13A is a configuration diagram for each class.

Classes are connected to each other through Object Properties. Since the User Class, which means user, is the subject of all situations, the Context Class and hasContext, which means the situation, the Deveice Class and hasDevice, which means the device, the Profile Class and hasProfile, which means the profile, the Disease class and hasDisease, which means the disease, and the service. Service Class which means hasClass has the relationship between hasClass, Motion class and hasMotion which means motion, Sensor Class which has sensor and hasSensor. 13B is a configuration diagram for each Object Properties.

Each class expresses each property through data property. The relationship between context and blood pressure is hasC_BloodPressure, and device and device type have hasKind. 13C is a part of a configuration diagram of Data Properties.

Each class has basic Individuals. This is the information that the class has, and FIG. 13D is a part of the configuration of Individuals.

Next, the service analysis module 35 searches the user's context information and ontology derived from the context engine 33 (or the inference engine 33b) to find a service suitable for the user. As a result, the service is provided to the user. The provided service is shown in FIG.

The inferred situation can be divided into 'daily or exercise' and 'emergency' as shown in FIG. 14 above, and provides a mobile cloud service through situation inference as shown in FIG. 15 according to each situation.

The service analysis module 35 requests emergency services to the hospital closest to the user's location when the reasoning result is 'emergency', and provides the records to the corresponding hospital and paramedics if the medical records of the user are needed.

Next, the effects of the present invention will be described in more detail with reference to FIGS. 16 to 19.

In the following, the purpose of performance evaluation and comparative analysis is to prove that the system of the present invention provides a more suitable and efficient recommendation service by using various user's situation and profile information than the existing situation recognition systems. Experimental evaluation of the present invention was carried out with an emphasis on whether to provide a recommendation service to the user appropriately and quickly in consideration of the situation information that was not properly processed in the existing system. The experimental results were compared with the existing system for evaluation.

The purpose of the system of the present invention is to provide a service suitable for the user by using not only the situation in which the user is located but also the information of the profile having the history of the personal information and the situation information of the user. To implement the system of the present invention, the middleware running in the window environment and the application running in the mobile device are implemented according to the framework of the proposed framework. Through the actual implementation of the system of the present invention, experiments were conducted on how suitable, efficient, and quicker recommendation and provision of services required by a user were performed. In addition, in order to generate and process a situation in a mobile cloud environment, real sensing information should be collected from a sensor device.

The system of the present invention provides a more efficient and suitable recommendation service suitable for the user's situation by inferring the situation using the profile information containing the user's situation information and personal information in the virtual environment.

The situation recognition application scenario for the system of the present invention is as follows. The reference value of the sensor data for each situation according to the health state of the user is shown in FIG. 16.

Examples of each situation of User 1, User 2, and User 3 are as follows.

Case 1: When user 1, user 2, and user 3 have all 110/150 as blood pressure sensor value,

-User 1 is a person with hypertension who has a normal blood pressure of 100 / maximum blood pressure 140 and does not correspond to 'emergency' criteria of 'lowest blood pressure less than 80 or maximum blood pressure greater than 160'. It does not provide a service because it is considered to be an exercise.

-User 2 is a user who shows the lowest blood pressure 80 / maximum blood pressure 120 and corresponds to the 'emergency' criterion 'lowest blood pressure is less than 60 or maximum blood pressure is greater than 140'. Provide emergency services / recommended hospitals' services.

-User 3 has the same result as User 2.

Case 2: When user 1, user 2, and user 3 have all entered the temperature sensor value of 36.8 ℃,

-User 1 does not correspond to the case of less than 35.7 ℃ or greater than 37.2 ℃, which is the criterion of emergency, so it is judged as 'daily or exercise' and does not provide services.

-User 2 does not correspond to the case of less than 35.8 ℃ or greater than 37.3 ℃, which is the criterion of emergency, so it is judged as 'daily or exercise' and does not provide services.

-User 3 does not correspond to the case of less than 35.6 ℃ or greater than 37.1 ℃, which is an emergency standard, so it is judged as 'daily or exercise' and does not provide services.

Next, items for performance evaluation of the system of the present invention for providing a recommendation service using the user's context information and profile information are as follows.

** Accuracy and efficiency of service delivery

   -User service accuracy

-The situation at the time of service request and the appropriateness of service provided

To test and simulate the system of the present invention, we implemented middleware and mobile applications. Middleware runs on the Windows XP operating system, and mobile applications are built on the Android platform.

A screen modeling the situation proposed by the present invention using Prot ㅹ g # 2000 is shown in FIG. 17A. FIG. 17B is a screen modeling sensors, FIG. 17C is a screen modeling disease data, and FIG. 17D is a screen modeling a profile.

18A, 18B, and 18C are user interfaces implemented in the Android platform. 18A and 18B show the contents of a user profile currently having hypertension and heart disease, respectively. The user of FIG. 18C shows a good state of health without any history as can be seen in the profile.

FIG. 18D is a screen showing the location of a hospital close to the user's location information when the user's situation is inferred using the situation information and the profile.

18E is a screen for embodying a medical record of a user provided to an emergency worker or a nearby hospital when an 'emergency' situation occurs. The medical records of these users are located in each hospital's medical care system and can be viewed as needed through the mobile cloud.

19 shows the result of comparative analysis between the system of the present invention and the existing frameworks. Context Toolkit, CoBrA, Gaia, and SOCAM, which are existing frameworks, and the system of the present invention were classified and analyzed into context models, context processing, mobile service support, service discovery, and history context information.

As shown in FIG. 19, the framework using the ontology-based OWL in context processing is CoBrA, SOCAM, the system of the present invention, and the framework supporting the mobile service is only the SOCAM and the system of the present invention. In addition, the framework using the inference engine in service discovery is SOCAM, and the framework using the knowledge base with the inference engine is CoBrA and the system of the present invention. It can be seen that the framework using the profile considering not only the user's history context information but also the preference or propensity as the history context information is only the system of the present invention.

The system of the present invention utilizes profile information together with the user's situation information, so that a more accurate service can be recommended and provided, while data corresponding to personal information of the user's profile information has to be input in advance. .

As mentioned above, although the invention made by this inventor was demonstrated concretely according to the said Example, this invention is not limited to the said Example, Of course, a various change is possible in the range which does not deviate from the summary.

10: sensor 11: mobile
20: cloud environment 30: situational awareness system
40: cloud database

Claims (9)

In the active situation recognition system for the recommendation service in the mobile cloud environment,
A sensor processor configured to receive raw data collected from the sensors;
A situation information collector for generating situation information by converting the raw data into a data format defined by an ontology;
A situation engine for performing redundancy check on the situation information and performing inference using the situation information and user profile information;
A knowledge base manager for managing the situation information and the user's profile information ontology; And
Active situation recognition system for a recommendation service in a mobile cloud environment, characterized in that it comprises a service analysis module to find the service for the user by searching the context information and ontology of the user derived from the context engine.
According to claim 1, The situation information collector,
A stream processor for converting the raw data into digital; And
Active situation recognition system for the recommendation service in the mobile cloud environment, characterized in that it comprises a sensor data converter for converting the digitally converted data into a data format defined by the ontology.
The method of claim 2, wherein the situation engine,
A valid processor for checking redundancy of the situation information; And
Active situation recognition system for a recommendation service in a mobile cloud environment, characterized in that it comprises an inference engine to infer using the situation information and the user profile information passed through the effective processor.
The method of claim 3,
The effective processor is an active situation recognition system for a recommendation service in a mobile cloud environment, characterized in that for checking redundancy using the tolerance (hereinafter offset) of the context information.
The method of claim 3,
And the valid processor discards the contextual information when it is determined that the contextual information is duplicated.
The method of claim 3,
The reasoning engine is an active situation recognition system for a recommendation service in a mobile cloud environment, characterized in that using an axiomatic rule-based reasoning system.
The method according to claim 6,
The rule used in the reasoning engine is an active situation recognition system for a recommendation service in a mobile cloud environment, characterized in that defined by Rule Markup Language (RuleML).
The method of claim 3, wherein the knowledge base manager,
A profile handler managing user profile information; And
Active situation recognition system for the recommendation service in the mobile cloud environment, characterized in that it comprises an ontology handler for managing the context information ontology-based.
The method of claim 1,
The situation information is active situation recognition system for the recommended service in the mobile cloud environment, characterized in that any one or more of the pulse, body temperature, blood pressure.

Images