KR20180009408A - Placeness Extraction and Inference System based on Social Network Analysis and Ontology - Google Patents

Abstract

The present invention relates to a placeness extraction system using social network analysis and ontology, and a method thereof. The method comprises: collecting unstructured data on the Internet such as social network service (SNS) and blogs; generating structured data based on ontology containing five ws and one h; and analyzing and inferring physical, social, and cultural meanings of a place based on the data.

Description

FIELD OF THE INVENTION [0001] The present invention relates to a social network analysis and ontology,

The following explains how to collect unstructured data on the Internet, such as social network service (SNS) and blogs, to generate formalized data based on the ontology that includes the two-way and five-way rules, And a method and system for analyzing and reasoning cultural meaning.

Recently, as the penetration rate of mobile devices such as smart phones has greatly increased, the development of Location-based Service (LBS), which searches and recommends peripheral information based on user's real-time location using GPS built in mobile device It appears to be prominent. Typical examples are Yelp, Foursquare, and Google Now. However, existing location-based services perform search and recommendation based on the user's location and the physical characteristics and superficial information of the place to be searched, and do not reflect the situation of the user using the service and the specific intention of the user It has limitations. Although there are systems such as exposing the ratings and opinions of other users together in the search results and reflecting the opinions of friends in the SNS, the social diversity and intention of the users who are still recommended, and the context of the situation at the time of recommendation And there is a problem that accuracy of searching and recommendation of information is lowered because it is omitted in the recommendation process.

According to Dourish and Harrison's research, one physical space can be interpreted in various meanings by the social context of the users who visit the space, and thus, placeness). Current location-based services do not take into account the social context or the characteristics of visitors to the sites they are searching for. Therefore, in order to search for and recommend accurate places, it is necessary to determine whether a specific place meets user's intention and social context The need for manual service providers to collect clues and provide personalized information.

Conventional technology "Behavioral ontology model and schema design using contents of social network service" proposes behavior ontology model and schema that can analyze behaviors posted on SNS and express human behavior. However, since the existing inventions analyze only the behavior of the user and use it for information recommendation, there is a problem that information about the user's context is lacking. It is difficult for users to understand the intention and meaning of a certain act by attempting to understand only the behavior of the user without considering the context.

The present invention relates to a method and system for locating a place in a specific place based on social data based on an instance of a visitor (Dweller), a companion (With), a What, an activity, an Opinion, a space, , Faceted information accumulation in a place by modeling it in terms of attributes, attributes, and relationships, and infer the physical characteristics, cultural meanings, and social behaviors of the place based on the ontology knowledge system, To provide precise search and location information.

The data analysis module automatically collects and integrates feature keywords in response to various types of web services such as social network services and blogs. The data abstraction module annotates the extracted keywords, The inference engine accumulates the place visit experience by a set of standardized keywords. The inference engine generates a decision tree based on the place-based ontology from the formalization data, and derives the different place names based on the frequency number and provides them as an interface .

The present invention relates to a method and system for locating a place in a specific place based on social data based on an instance of a visitor (Dweller), a companion (With), a What, an activity, an Opinion, a space, , Faceted information accumulation in a place by modeling it in terms of attributes, attributes, and relationships, and infer the physical characteristics, cultural meanings, and social behaviors of the place based on the ontology knowledge system, To provide accurate search and location information.

BRIEF DESCRIPTION OF DRAWINGS FIG. 1 is a diagram showing a configuration diagram of a locality extraction and shaping system. FIG.
2 is a diagram showing a detailed configuration diagram of a locality extraction and formatting system.
Figure 3 is a plot of the raw data used for locality definition by Cranshaw et al.
Figure 4 is an example of processing user input data in a study by Shaefer et al.
5 is a diagram illustrating a system architecture and process for extracting pairs of verbs and nouns from the community production period.
6 is a diagram showing a structure of a database.
7 is an example of structuring a place-related element based on an ontology in a location-based service.
Figure 8 is an example of an ontology and instance used in the study of Alaya et al.
Figure 9 is an example of a hierarchical definition of a place.
FIG. 10 is a diagram for explaining Placeness Structualization and Placeness Inferencing steps. FIG.
11 is a diagram showing a locality-based ontology model.
12 is a diagram showing the results of the image analysis test.
FIG. 13 is a diagram illustrating requirements analysis and locality-based formalization and reasoning techniques for deriving the multisocial meaning of space.
FIG. 14 is a diagram for explaining a detailed feature location-based property shaping and reasoning technique based on a big data analysis for real-time context application and place segmentation in a real-time place recommendation.
15 is a diagram for explaining a place-user multi-dimensional network related relevance inference technique for reflecting a user's instantaneous / immediate context.
FIG. 16 is a diagram for explaining a technique for shaping emotional characteristics of on-line related information for recommending a preferred place in consideration of user's emotions.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings.

The present invention relates to a space-based data mining core technology for supporting an optimal real-time location-based intelligent information recommendation service suitable for a user's situation in a smart space capable of acquiring various information from a smart phone and IOT, This course analyzes the requirements for placeformatting and reasoning techniques to derive multisocial semantics of space based on behavioral history and describes elemental technology design.

When a person visits a place, the place is determined by combining the purpose of the visit, the current location, the persons accompanying the person, and other individual requirements. Even in the same place, it is recognized differently depending on the characteristics (taste, age range, etc.) of the individual. In this way, we define the multifaceted characteristics of one physical space by the social activities that take place in that space as the locality. Considering this locality is essential information in order to improve recommendation accuracy when providing real-time intelligent information based on location, but today's situational information-based location recommendation service takes full consideration of the key elements needed to infer placeality There is a limit that can not be done. In order to solve this problem, the present invention analyzes elements required for inferring locality and requirements for inference method for intelligent information recommendation, and designs locality-based formalization and inference element technology to achieve requirements.

- Place inferences and application techniques

With the increase of smart phones, users can search anywhere and anytime regardless of location. In the same vein, the emergence of location-based services suggested a new paradigm of location search by providing users with the most useful information at their current location. In architecture, specific information related to the place is called 'locality', and attempts to define 'information' and further 'meaning' in a specific space have been studied mainly in (1) public places or (2) smart home domains. For example, users' behaviors in urban and public places can be guessed through user feedback, GPS, etc. to define characteristics of a specific location, and IoT devices and sensors that interact with a smart home Identity.

The ultimate goal of defining locality is to get physical or static information and contextual information about a specific location by a person or machine. It is necessary to have a formalized grammar system so that information related to a place can be easily retrieved and utilized Do. According to previous studies, a representative knowledge system for a place has been used 1) a relationship structure 2) a keyword / tag 3) a raw form.

- Approach that defines raw data as locality

The raw form of locality definition is a way of utilizing data from sensor or user input without any filtering / matching. Unprocessed data itself is not suitable for place classification, but has the advantage of high statistical significance and a high degree of personal context of user input.

- Define the place where the relationship structure is applied

Relational structured form is the best way to express the contextual meaning of properties that can have placefulness by expressing correlation and causality of data in graph or tree structure. The way in which the ontology is designed and the relationship is structured is typical, and a circular or hierarchical structure can be utilized.

Issue 1: Missing key elements for locality inference

Although the studies that define the place property as the relation of the data are very suitable for the classification and the search of the place, the limited complexity of the relation and the difficulty of extracting the related data have structured only a limited number of categories such as subject-verb and put them into the ontology model . Therefore, if we define only the definite elements such as subject-verb rather than all the static and contextual information about the place, there is a limitation that we can not find the minute difference between the places or record the little details of the place.

Issue 2: Multi-faceted place to space is impossible to express.

The concept of locality-based definition of place-based services and the concept of locality previously studied has a 1: 1 relationship between location and locality. Therefore, it is theoretically impossible for a place to have two place properties. In addition, if a keyword tag is used, it is impossible to match even a very similar keyword. Therefore, a 1: N property There is an issue that it can not grasp precisely.

Issue 3: Excessive user intervention for locality reasoning occurs.

In the services and researches described above, various attempts have been made to extract the place where the user has experienced and felt toward a specific place, but the extraction method is designed in a way that is inconvenient to the user. For example, there is a problem that continuously correct input for the purpose of extracting information or always utilizing the smartphone's sensor is interfering with the user, consuming the battery, consuming security and infringing privacy .

Based on the above - mentioned issues, we analyzed the requirements of location - based data mining core technology for real - time intelligent information recommendation service in smart space.

Based on the physical characteristics of space and user 's history of social behavior, we propose a place - based stereotyping and reasoning technique to derive the multisocial meaning of space.

SR101: (It is necessary to apply the factor that corresponds to the two or three principles to place inferring.) In order to express one place by taking into account the social characteristics of people and the physical characteristics of space, (5W1H; Who, When, Where, What, How, and Why) to consider physical locations together. "Who" can describe social relationships with individuals and companions, "when" does not only mean absolute time but also the various meanings of the time (season, holiday, morning / Afternoon division, etc.). "Where" can express unchanging physical attributes such as space name, classification, and location. "What" is an element that plays a very important role in interpreting the same space in various social meanings in this study, and it should be able to express what actions users performed. Since "how" acts as an adjective and adverb that modifies what has been described above, this helps to distinguish the atmosphere or the nature of the user's preference, even for the same purpose. The factor of "why" is related to the characteristics of the order of movement from one place to another.

SR102: (The expression of mood and situation should be used in place reasoning as a detailed factor.) The reason for users to visit and work in a certain preferred place is to achieve the purpose first, If there are several places of the same kind (assuming that there is no distance constraint), then choosing one of them is influenced by various other details of the place. These sub-elements can be roughly describing the atmosphere of the space, or expressing the behavior and space of the user. Therefore, it is necessary to collect the adjectives / adverbs that express the atmosphere or the subdomain from the visiting history of users and use them as detailed elements.

SR103: (Definition of the semantic relationship between the two or three elements.) As people make decisions based on the two-way and five-way rules in the everyday domain and consider factors that have important influence among them, Once the site visit history is represented by factors such as the two-way hierarchy and atmosphere, the computing system must define associations between elements so that they can identify important (meaningful) locations from a combination of these elements. For example, it is defined that the relationship between "location" and "activity" arises between the "restaurant" where "VIP treatment" or "celebrity" occurs mainly in the same restaurant and the restaurant that visits " Need to be. Also, the association should be defined between "user" and "activity" in order to know which users mainly visit certain spaces and perform certain actions. In this way, it is necessary to define the relation between elements necessary for place reasoning and use it as a knowledge system. Therefore, ontology should be utilized as a tool to express knowledge system flexibly and extensively.

SR104: (It is necessary to be able to extract suitable places for some element combinations.) Once the association between the elements required for locality inference is defined, it is possible to grasp the elements that each site mainly has from the site visit history data. Therefore, it is possible to combine key elements to find out which space best matches those elements. This goes hand in hand with the process of reasoning that people routinely perform when deciding where to visit. Therefore, the location reasoning system should not only identify the main elements of each place from the visit history data, but also be able to find relevant places by using the main elements.

SR105: (It should be able to express multi-locality with respect to one place.) If the concept of placeality proposed in the previous study can express only 1: N relation, It should be comprehensive.

SR106: (The user intervention should be small when inferring the place locality.) The locality extraction method proposed in the present invention should be non-invasive to users and secure in terms of security and privacy.

We propose place formality and reasoning technique.

- Block design

Placeness of a space can be defined as the accumulation of all the social actions of the people who are accomplished there. In recent place-based studies, a similar concept is introduced: 1) raw (raw) ) Data to handle low level locality, 2) to analyze macroscopic locality by matching predefined keywords in place documents, or 3) to use systematic And to try to understand the correlation. However, it is possible to embrace the multifaceted nature of the place at once, but before it expresses the macroscopic and detailed locality at the same time, The structure of research designs is limited. Therefore, the present invention proposes 1) a relational structure to overcome the limitation of existing place-based stereotyping methodology, that is, an ontology and 2) an inference module for inference, verification, and application of attributes of the proposed ontology from data.

The proposed system is designed in two parts: 1) site stereotyping and 2) location inference. The formalization part is a part for designing classes and features to be included in the ontology, and the reasoning part includes a data mining and feature extraction method for collecting the characteristics of the ontology defined above. The two parts are closely related, and each result can be used as an input.

The proposed system defines the definition of the initial ontology based on the ontology obtained empirically through the existing research or data analysis results. Then we mine the data needed for the attribute and get the relevant information from the raw data through the Inference module. The obtained attributes can be used for redefinition or improvement of the formalization method through module testing and ontology verification.

- Place sex stylization

The preliminary research is the first step to establish the method and criteria for the gender stereotyping. The concept of the place of gender has been largely classified into 1) a private space such as a smart home, 2) public) I found that public places exist in two domains. Since the locality proposed in the present invention should be applied universally, not on one domain, the ontology is designed without any domain.

The present invention is based on the fact that attributes that all the place studies in common are important can be categorized into so-called two-way rules. The present invention is based on the fact that all attributes of place property are classified into five classes including time, visitor, space, property.

In order to enable multidimensional locality expression for one physical space, the present invention is designed as a place-based stereotyping ontology as shown below. Although there is only one physical attribute of space, it can be varied in various ways as to what social activity it matches. In order to take into account these multidimensional characteristics, we have defined the optional relationships between classes and attributes. For example, a space of a superclass may have a relationship of visits with another parent Dweller, or may have an association of isIn with an activity of another superclass.

The specification of the subclass defined for each superclass is as follows.

Time: The concept of time in the place name should not only consider the time when the visitor arrived at the place but also include how long the visitor stayed and whether the visit was a weekday or a holiday, Do. For example, touring places such as sightseeing spots and amusement parks have much more visitors on holidays and weekends than on weekdays, so determining holidays is an important factor in choosing a place. Therefore, the Time class is defined to have subclasses such as DepartureTime, Duration, DayofWeek, and isHoliday as well as Timestamp, which means visit time.

Space: In the present invention, the concepts of space and place are separated in order to effectively express place properties. A space is a section divided physically according to purpose and function, and a place is defined as a semantic by various activities occurring in the space. Therefore, in the proposed ontology, the spatial class refers only to physical characteristics. Here, in addition to the basic information such as address, latitude / longitude, weather, entropy (how various places are located around), traffic information and population density .

Dweller: Visitors' information is an important indicator of place locality definition. This is because even the same space can be used for different purposes due to individual factors such as the difference in visitor's social characteristics, the size of the visitor group, and the taste of the visitor. However, there is a difficulty in collecting personal information compared to the importance of information of visitors. Therefore, the proposed ontology defines the Dweller class as the most basic information such as the number of visitors, visitor entropy (diversity), sex, and age, which are collectable or inferable demographic factors.

Activity: Since the definition of place-property itself is the accumulation of the social act of people who take place in place, Activity has pivotal properties in place-based ontology. Activity is a class that can be inferred from various data (picture, text, hashtag, etc.), and the properties are also varied. By including the previous activity, such as LastActivity, the relationship of the successive activities can be included.

Opinion: Although the opinions obtained by the visitors about the place are not yet covered in the preliminary study, the related properties are currently defined, assuming that the related information can be deduced as a set of adjectives and adverbs existing in the text.

Since the ontology initially proposed is an ontology based solely on preliminary research, it will be redefined and refined according to the results of the verification process based on the actual collected data and the results of the place reasoning work.

- Place sex reasoning

Social media is often used to study sociology, organizational behavior, or network in that it captures users' behavior and psychology in an implied form. Among them, It is very useful in that the state can be put together in real time. In social media, there are many kinds of data to be used for each service. In the present invention, data used in experiments are 1.3 million images, hash tags, and text in the Instagram.

In order to deduce the place from the image, we used the object recognition algorithm using Deep Neural Network, which has been studied most recently in artificial intelligence field.

The present invention develops requirement analysis and data formatting techniques for developing core technology based on locality-based data mining. To develop requirements analysis and place formatting and reasoning technology to derive multisocial meaning of space. Develop data structure formalization and reasoning techniques for inference of detailed place characteristics. A place for screening of the most popular places suitable for the user context - Multi-dimensional network connection among users also requires reasoning. Analyze and develop modeling techniques. Develop emotional characterization of online information related to detailed place for recommending preferences considering user 's emotion.

- Requirements analysis and place formatting and reasoning technology development to derive multisocial meaning of space

We analyze location reasoning requirements based on physical characteristics of space and user's social behavior history, develop locality-based ontology design and ontology-based place-based data formatting technology, develop place-based reasoning technology of formalized data base do.

To develop requirements analysis and place formatting and reasoning technology to derive multisocial meaning of space. Based on the survey of items used in place recommendation in existing location-based services, we define the key elements that constitute the locality of the 5W1H-centered hierarchy.

We develop the ontology ontology design and the ontology based place data formatting technology. We design a place-based ontology that defines the relationship between social behavior, spatial characteristics, and social context, and develops the locality-based data formatting technique for multidimensional context information based on the association defined in the ontology ontology.

Develop place-based reasoning techniques based on formalized data-based data. We develop a place inference technique that utilizes a Bayesian network - based probability model.

We investigate the place formatting method of domain by domain. Especially, based on location - related research, methodology and basis of place - In the existing locality study, the place is defined as raw data, keyword or relation structure, but 1) raw data is not utilized as context, 2) matching of simple keywords can not distinguish delicate feature difference of place, and 3) There is a limitation that it is impossible to express multi - dimensional place in space.

We define place - based ontology to overcome limitations and problems of existing research. Based on the surveyed contents, it is possible to design an ontology with five classes and 44 independent properties of Time, Space, Dweller, Activity, Opinion and to consider and express the characteristics of multidimensional locality by reflecting the probability model. Defines an attribute.

We perform place inference based on designed ontology. We implement social media mining, image processing, natural language processing, and context inference modules to understand non-invasive social actions and contexts in place.

In addition, it verifies the suitability of the placeable ontology attributes designed through the initial data mining data and the locality inference algorithm, and additionally reflects the inferable place property, and combines social media with OpenAPI to access the place and public data defined by individual contexts Unify the place where possible.

- Detailed data analysis based on big data analysis to apply real-time context and place subdivision in real-time place recommendation.

We analyze the technical requirements for inferring the characteristics of each of the most popular places in the upper space, and develop a technology to formulate the data structure for inference of detailed place characteristics of various data bases such as SNS, traffic, and social statistics.

It analyzes the technical requirements for inferring the characteristics of each detail popular place in the upper space, defines the relation between the upper place and the detail place through the detailed place characteristics and distribution, Define a specifiable locality list through a data set investigation and each data.

SNS, traffic, and social statistical data, etc. We study place specific reasoning that can be obtained from individual data sets and study place specific reasoning model which shows high performance in user - place matching by integrating various data.

- Investigate availability and extractable characteristics for possible data sources.

Micro Blogs (Twitter, Instagram) is short, but it is likely to be used in real time because it reflects a lot of references. Blogs reflect real-time situations, but they contain detailed and detailed information. Public data such as industry distribution, population, and other public data represent long-term characteristics of each region, and data such as traffic volume and weather can reflect real-time characteristics. Finally, Instagram and public data are selected for detailed site specific extraction.

- Design the keyword-to-keyword clustering for feature extraction.

Place Gender Inference Select elements for inference of detailed place characteristics according to the ontology derived from the detailed task and classify them into word sets. The elements representing the detailed place characteristics are systematized on the basis of pre-registered relationships (meaningfulness, semantics, etc.) and reworked into a behavior suitable for reasoning.

- Provide a model of relational reasoning between the top place and the detailed place characteristics.

Previous studies have used placeability without taking into account the characteristics, distribution, etc. of detailed sites. We design a connection model between the detailed place characteristics and the top place locality based on the number and distribution of detailed places.

Furthermore, we will continue to add a set of words for extracting features of the detailed place, and implement the ontology of the relationship between the top and bottom places.

- Place for user's immediate / immediate context reflection - Multidimensional network related degree inference technique between users

A multi-dimensional network construction methodology for defining location-user context relationship. Analyzes requirements and develops connection pattern analysis and modeling technology for relevance inference based on multi-dimensional network between users and users.

Analyze requirements of multidimensional network construction methodology for defining location-user context relationship. We define place-sex and user-context-related elements and their relationships for defining a place-user context relationship, and investigate a multidimensional network construction methodology that can effectively represent the relationships between place-related and user contextual elements and their relationships.

We will develop modeling technology through connection pattern analysis for relevance inference based on multi-dimensional network among users. A multi-dimensional network in which the relationship between location corpus and user usage history data is defined quantitatively and qualitatively is constructed, and analysis of the multi-dimensional network characteristics between users and users is performed through analysis of connection pattern between users and users. Research.

Existing research place recommendation methods investigate existing user context elements and place visiting models (popularity-based, collaborative filtering, etc.) that are used for place recommendation. In the previous research, the place is recommended to the users by using the frequency of visiting places or the filtering of collaborations, but it does not consider the actual place visiting pattern to select the appropriate place property according to the user's context.

We define place-user relationships through multidimensional networks to overcome limitations and problems of existing research. In order to reflect the actual place visiting pattern to the place recommendation algorithm, the relationship between the place and space, space and tag is defined, and each relation is also defined as a multidimensional network according to the detailed place characteristics and emotions.

Develop a place-user related inference technique for choosing an appropriate place for user context. We define the visit time, motivation, and purpose as the user context elements related to the site performance for the recommendation of the place according to the user context, and use the existing place recommendation (popularity, user similarity, Verify the implementation and performance of the study.

In addition, we develop a multi-dimensional network building module between user and user defined above, design and develop related inference technique in multi-dimensional network for multi-dimensional network for user's context selection, Evaluate network related relevance techniques.

- Describe the emotional characteristics shaping technology of the online information related to the detailed place for recommending the preferred place considering user's emotions.

We develop emotional analysis technology specialized for place-sex social data and develop technology of place-based based mass sensitivity stereotyping. Develop emotional analysis technology specialized for place-social data.

In the preprocessing stage, morphological analysis is performed to exclude parts of speech that do not reveal emotion (eg, nouns), and devise an extension technique to extract emotional expressions that are actually used in unstructured social data.

- Development of technology for stereotyping of public feeling based on place

Place identity is revealed in a collective form of public action for a specific place. The emotional vocabulary that appears as the action vocabularies is used to formalize the locality - based emotional characteristics.

- A related research on location-based emotional analysis

Existing approaches for location-based social data analysis have limitations in that they analyze emotions using the most common emotional words, not the words that real people use to express emotions in place-based social data. In addition, there are limitations in analyzing the emotions of the simplest dimension such as positive / negative about the place of the unit (eg, state, county) that is difficult to be the object of the recommendation although studies are conducted to understand the public emotion.

- In order to compensate the limitation of the existing research, we designed the technology for the emotional analysis of the place-specific social data. We designed the extension method based on Pointwise Mutual Information (PMI) in order to express emotions with non - common words such as fluency, ota, and new word in place - sex social data and to increase the coverage of emotional expression that is actually used. For effective emotional analysis, parts of speech (eg, nouns) that do not reveal emotion in the preprocessing step are excluded based on POS tagging.

- Technology for designing public sensibility based on place sex

The place name is revealed through action and it should be formalized in consideration of the public feeling. In the place-based social data, it is assumed that the frequently used words in the appropriate level place (for example, restaurant) Design a word embedding module to learn.

It is possible to understand the words of vector form obtained by using Word2Vec algorithm by clustering using distance based clustering algorithm and to recognize adjective words expressing emotion in close proximity to the action vocabulary as emotion related to each action. Formalize the castle-based public emotion.

Furthermore, we implement emotional score calculation method that assigns weights based on user's context. Establish a place-specific vocabulary and enable effective analysis at high speed.

The apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the apparatus and components described in the embodiments may be implemented within a computer system, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA) , A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device , Or may be permanently or temporarily embodied in a transmitted signal wave. The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI > or equivalents, even if it is replaced or replaced.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (4)

1. A method for analyzing data related to a social network service in an Internet space and a location on a blog,
Collecting heterogeneous atypical content data collectively;
Inferring a keyword for a place by Dweller, With, What, Activity, Opinion, Space, and Time; And
Dweller, With, What, Activity, Opinion, Space, When
Further comprising the steps of: The method according to claim 1,
Collecting heterogeneous place data from social network services and blogs using a customized crawler periodically
Further comprising the steps of: The method according to claim 1,
Define an ontology that defines associations between keywords that play an important role in place-keeping, focusing on Dweller, With, What, Activity, Opinion, Space, and When
Further comprising the steps of: 4. The method according to any one of claims 1 to 3,
Dweller, With, What, Activity, Opinion, Space, and When extract keywords related to locality from unstructured data based on ontology and store them in database in a formalized format
Further comprising the steps of:

Images