KR102204992B1 - Big Data Analyzing System and the Method based on the User Profile Analysis - Google Patents

Abstract

The present invention relates to a system and method for analyzing and summarizing big data based on a profile of a data creator. The present invention automatically collects profile information such as the age, gender, residential area, and field of interest of the author who created the data, and analyzes the data using the data, so that it is possible to analyze and provide results for objects desired by users in various aspects.

Description

Big Data Analyzing System and the Method based on the User Profile Analysis}

The present invention relates to a technology for extracting/retrieving desired information from Internet text or images, and specifically, a big data analysis system based on user profile analysis that extracts information desired by a system user based on text and profile information written on the Internet. And to the method.

The Internet and various social media contain huge and diverse information called big data. Various studies are being conducted to extract and analyze meaningful information from such data, and one of them is a study that collects/analyzes and utilizes various opinion information described in big data. Examples of such studies include a study that analyzes the strengths/disadvantages of a specific product, analyzes the positive/negative reactions to the product, and summarizes what positive/negative opinions about the product appear on the Internet. In this study, as a result, positive/negative opinions on specific products are summarized and presented as follows.

● XX camera

■ Sharpness

◆ Positive Opinion: 142

-XX camera can be said to be the best in sharpness. Better than any other camera.

-Among the various functions of XX camera, sharpness overwhelms other cameras.

◆ Disagreements: 12

-I regret buying a blurry photo, XX camera.

■ Battery

◆ Positive comments: 52

-XX Camera has a really long battery life. It lasts a month without charging.

-The battery lasts really long once you charge it.

◆ Disagreements: 12

-It is really hard to charge the battery.

-The battery is too expensive and it is burdensome to purchase additionally.

Various methods are used to present such information. In general, information is presented to the user through the following procedure.

1. Collect documents from review sites where opinions are stated on specific products

2. Apply emotion analysis technology to the document and extract positive/negative opinions

3. The extracted positive/negative opinions are classified by attributes (Aspect, sharpness in the above example, battery) and presented to the user as a result.

Through this procedure, system users can respond by grasping what opinions are being exchanged on the web about the subject to be investigated. However, in the case of the conventional technology, the information of the author of the data who wrote the opinion is omitted.

This is a clear limitation in view of the fact that the information of the person who presented the opinion in actual marketing, for example, the user's age, region, and gender, are important factors. For example, in the case of selling a product targeting young people, it is necessary to understand what points users in their 10s to 20s who target targets rather than other targets as their strengths and what they think as weaknesses, so that the product can be promoted/improved accordingly. I can.

Therefore, a method of automatically analyzing basic information such as the age, gender, residential area, and field of interest of the data creator, and to analyze opinions and public opinion on the object to be analyzed based on this, and a method to effectively present it will be required. .

In order to solve the above-described problem, the present invention analyzes the profiles of users who generate data by analyzing data on the Internet, analyzes data based on the analyzed user profile information, and a group of users who generate the data to be analyzed. An object of the present invention is to provide a big data analysis system and method based on user profile analysis that provides profile information.

The present invention includes a user profile analysis unit that collects data generated by users and analyzes the profiles of the users based on the collected data; A data analysis unit analyzing data generated by the users and users who generated the data using profile information of the users; And a result providing unit that provides the result of the analysis by the data analysis unit, wherein the result providing unit provides both the data analysis result by the data analysis unit and the profile information of the users who have generated the data. It provides a big data analysis system based on it.

According to an aspect of the present invention, the result providing unit provides profile information of the user who generated the data when clicking on the data analysis result or when a cursor is positioned on the data analysis result, a big data analysis system based on user profile analysis. Provides.

According to another aspect of the present invention, the result providing unit provides a big data analysis system based on user profile analysis, which provides the result of analyzing the profiles of users who generated the data when the data analysis result is selected.

According to another aspect of the present invention, the result providing unit provides a result of analyzing the profiles of users who generated data at the specific time when a specific time point is selected from the results analyzed over time. It provides a big data analysis system based on it.

According to another aspect of the present invention, the user profile analysis unit analyzes the profiles of the users using at least one of user information included in data, user information inferred from the data, and user information provided by a service provider. It provides a big data analysis system based on human user profile analysis.

According to another aspect of the present invention, when the user profile analysis unit collects data, a big data analysis system based on user profile analysis is provided to analyze the user's profile by collecting other data generated by the user who generated the data. to provide.

According to another aspect of the present invention, the user profile analysis unit provides a big data analysis system based on user profile analysis in which the user profile analysis unit analyzes the user's profile using user information provided from the site where the data is started when collecting data. do.

According to another aspect of the present invention, the data analysis unit limits the range of the users and analyzes the data generated by the users included in the range based on the profile of the users, a big data analysis system based on user profile analysis. Provides.

Meanwhile, the present invention includes the steps of collecting data generated by users; Analyzing the profiles of the users using the data; Analyzing data generated by the users based on the profiles of the users; And it may be used as a big data analysis method based on user profile analysis including the step of providing an analysis result of the data and profile information of the users.

The present invention provides a big data analysis method based on a data creator's profile, so as to determine how the trend of public opinion changes by age, gender, region, and interest field of a specific person/company/product/company/organization, etc. In the case of product marketing, it is possible to check what opinions are made on the main target target of the product, so that it can be selected as a major promotional element or improvement target, and what points are highlighted as positive factors by age group for a specific person. Find out if the dot stands out as a negative factor so you can cope with it.

In addition, the present invention makes it possible to analyze trends by age, gender, region, and field of interest when analyzing trends based on big data, thereby identifying the characteristics of the data creator group by specific issues and opinions, specific positive/negative/neutral It enables monitoring of users with opinions, identification of the characteristics of user groups with specific positive/negative/neutral opinions, user monitoring for specific trends/issues, and identification of the characteristics of user groups for specific trends/issues.

1 is a block diagram showing the structure of a big data analysis system based on user profile analysis according to an embodiment of the present invention.
2 is a flow chart showing a process of generating user profile information by a big data analysis system based on user profile analysis according to an embodiment of the present invention.
3 to 6 are diagrams showing an example in which a big data system based on user profile analysis according to an embodiment of the present invention provides an analysis result.

Advantages and features of the present invention, and a method of achieving them will become apparent with reference to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various forms different from each other, and only these embodiments make the disclosure of the present invention complete, and common knowledge in the technical field to which the present invention pertains. It is provided to fully inform the scope of the invention to the possessor, and the invention is defined by the scope of the claims.

Meanwhile, terms used in the present specification are for explaining embodiments and are not intended to limit the present invention. In this specification, the singular form also includes the plural form unless specifically stated in the phrase. As used herein, “comprises” and/or “comprising” refers to the presence of one or more other components, steps, actions and/or elements in the mentioned component, step, operation and/or element. Or does not preclude adding. Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

In the present invention, a user who describes various data such as opinions and information in big data is referred to as a data creator, and a user who uses the analysis result using the system is referred to as a system user.

The object of the present invention is largely divided into two. 1) The first purpose is to automatically analyze the basic information (user profile) of the data creator, and 2) the second is to effectively analyze/summarize/present big data based on the analyzed author's profile.

The basic information of the data creator can be obtained directly from the service provider through API, or through information extraction and inference. The method of obtaining information from a service provider is a method in which the data creator takes and uses the basic information of the data creator presented on the website where the data was created. In this case, it is possible to easily and accurately obtain the basic information of the author of the data, but there is a problem that the information is not provided by a large number of web sites. Therefore, in the present invention, by analyzing the profile picture of the data creator, the written article, the introductory article, and the social network, information such as the user's age, gender, residential area, and object of interest is extracted using information extraction technology or inferred through the written article. It tries to acquire the information and use it for big data analysis.

The present invention also proposes a method of utilizing the profile information of the data creator when analyzing/summarizing/presenting big data. For example, by limiting the data creator to be analyzed using basic information such as the age, gender, and area of residence of the data creator, and information such as the object of interest and frequently mentioned keywords, only the information of a specific level is analyzed/extracted/summarized or analyzed. It is also possible to grasp the characteristics of data creators who have described a specific part of the content. For example, if “authoritative” was selected as a negative opinion when analyzing an opinion on a specific person, it is possible to grasp the disposition of the people who described the corresponding opinion and the age range. Another example is a smartphone in his twenties. Interested data creators can also see what other interests are. Through this analysis method, more in-depth and various analysis is possible.

1 is a block diagram showing the structure of a big data analysis system (hereinafter, a big data analysis system) based on user profile analysis according to an embodiment of the present invention.

The big data analysis system includes a user profile analysis unit 100, a data analysis unit 110, and a result providing unit 120.

The user profile analysis unit 100 analyzes text or analyzes a user's profile based on information provided by a web provider.

The data creator profile can be identified by analyzing the source of the collected data and analyzing other data created by the data creator, or by extracting profile information provided by the site on which the data is posted from the source.

2 is a flowchart illustrating a process of extracting profile information of a data creator by the user profile analysis unit 100.

In order to extract the user's profile, the article written by the data creator is collected/extracted (S200). This can be done by using the characteristics of the site on which the user posted the data. For example, if the author of the data posted an article on his blog and the address of the blog is http://blog.XXX.com/XXX0025, http://blog. By collecting sub-web pages of the XXX.com/XXX0025 site, it is possible to collect articles written by the data creator. If the source of the data posted by the data creator is a social network such as Twitter or Facebook, other posts of the data creator who posted their opinions on the website may use APIs or web crawling technology provided by the site. To collect.

When the articles written by the data creator are collected/extracted, necessary information is collected from the collected web page using an information extraction technique (S210). The information extraction technique is an explicit information-based user information extraction method that extracts information of explicitly described data, an inference-based user information extraction method that extracts information through inference, and information provided by service providers such as meta information. There is a profile data analysis method that extracts user information by using.

The explicit information-based user information extraction method is a technology in which the data creator extracts information from the text that directly describes his or her profile information. For example, if there is the following data among the data created by the data creator, information such as the name, age, region, gender, and topic of interest of the user who posted the article can be extracted through the technology. For this, natural language processing techniques such as syntax analysis and relation extraction can be used.

I disclose my information.

name: XXX

Age: 24 (I am a flowery maiden.)

Where I Live: Yeouido

I like to swing dance , and I like to ride bikes and travel. ㅎㅎ There are many hobbies and interests.

But crucially, I don’t have anyone to share my hobbies with me.

The inference-based user information extraction technique is a technique that analyzes the data created by the data creator and infers the profile of the data creator in consideration of frequently used keywords, topics of interest, style, and social networks. For example, if the keywords frequently used by data creators are "flowers, cosmetics, bags, clothes, beauty, fashion, women's age", the commenter can be inferred that the commenters are women. In addition, if the commentator frequently uses vocabulary such as "Hull, Solkamal, Berkachung, Geumsapa", it can be inferred that the commentator is a teenager.

If the source where the opinion is posted is a social network, the age of the user can be inferred through network analysis. For example, if the majority of users on the social network are in their 20s, it can be inferred that the user's age is also in their 20s. In addition, information such as gender/age can be extracted by analyzing the profile picture of the commenter. User interests can also be inferred by analyzing data. If a lot of game-related vocabulary such as "Lineage, PlayStation, Game, Final Fantasy, Anipang" appears in the articles written by the data creator, it can be inferred that the game is the field of interest of the data creator. For such inference, classification techniques using various machine learning such as Maximum Entropy, Bayesian Learning, Support Vector Machine, and KNN algorithm can be used.

According to the inference-based user information extraction technique, not only objective information such as age, gender, and residential area of users, but also information representing the user's characteristics such as interests and dispositions can be extracted.

In the case of a specific service, profile information, etc., is provided separately, or provided in the form of an API by a service provider, or information is provided in a table form for easy extraction. For example, in the case of Twitter, which is an SNS service, the location of the place where the article was written can be provided and the user's residential area can be inferred using this. The profile data analysis method is a method of analyzing the commenter's profile data using the user's information provided by the service provider.

The user profile analysis unit 100 extracts a user profile from the posting of the data creator through the above-described methods and generates user profile information (S220).

The data analysis unit 110 uses the user profile provided by the user profile analysis unit 100 to perform analysis tasks such as web trend analysis and opinion analysis.

FIG. 3 shows an example of a method of searching for positive/negative/neutral summary information for a specific target by constraining it based on user profile information among methods of analyzing big data using user profile information.

Users can select a desired target group by restricting the personal information of publishers who posted comments on the Internet and analyze only the opinion information of the group. Various personal information such as gender, age, region, area of interest, name, photo, etc. can be used to limit the subject.

Such restriction information can be input in various forms. In the case of age, you can search by specifying the lowest and highest age among the target age groups (20-45) as search conditions, or you can search by selecting a specific age group (teens, 20s). For gender, you can select one of "male/female/all". In the case of an area, a map can be displayed on the UI and selected by clicking on a specific area in the map UI, or the desired area can be searched and selected by entering the area name in the search box. In the case of a field of interest, a specific field of interest can be selected using a browsing method or a search box. For example, a user may directly input an interest field of “smart phone” into a search box to give constraints, or there may be a method of selecting and selecting a selectable interest field list displayed in the UI by a browsing method.

The result providing unit 120 presents the analysis result to the user based on the content analyzed by the data analysis unit 110.

4 shows an example of a UI in which the big data analysis system presents information of a data creator who has created specific data.

When clicking specific data from various big data such as trend analysis and sentiment analysis or placing the cursor on the data, the data creator's age, residence area, ID, name, interests, keywords of interest, other articles written by the data creator, data You can check information such as other articles written by the author on a specific subject.

5 shows an example of a UI that shows analysis information on profiles of data creators with respect to summary information analyzed by the big data analysis system.

When providing positive/negative/neutral summary information for a specific target, group user profile information such as average age, gender, region, interest field, and interest keywords of users who wrote articles corresponding to a specific classification (earphones in FIG. 4) to provide. In the example shown in FIG. 5, approximate information (average age, etc.) of data creators who posted opinions on the earphones of the XX phone 5 can be checked.

6 is an average age, gender, region, interests of data creators describing data on issues (trends) on a specific date when the big data analysis system analyzes issues or trends based on time series using big data. This is an example of a UI that provides group user profile information such as keywords of interest.

In the example illustrated in FIG. 6, information such as average age and gender distribution of authors who describe data on the issue of release can be checked.

The above description is merely illustrative of the technical idea of the present invention, and those of ordinary skill in the art to which the present invention pertains, various modifications and variations can be made without departing from the essential characteristics of the present invention. Therefore, the embodiments expressed in the present invention are not intended to limit the technical idea of the present invention, but to explain, and the scope of the present invention is not limited by these embodiments. The scope of protection of the present invention should be interpreted by the claims below, and all technical ideas equivalent to or within the scope of the present invention should be interpreted as being included in the scope of the present invention.

Claims (17)

Collects online data generated by Internet users, extracts profiles of users who generated the data based on the collected data, creates individual profile information for each user, and adds other data created by each user. A user profile analysis unit that collects and extracts user information through inference based on a plurality of data of the same author and generates profile information of each user who has created the collected data;
A data analysis unit that extracts data generated by users meeting a specific search condition using the individual profile information and performs at least one of trend analysis and public opinion analysis on a limited user group as an analysis target; And
A result providing unit that provides the data analysis result analyzed by the data analysis unit and provides group user profile information for users who generated the analyzed data
Big data analysis system based on user profile analysis comprising a.
The method of claim 1, wherein the result providing unit
Providing the data analysis result by the data analysis unit and the profile information of the users who generated the data together
Big data analysis system based on human user profile analysis.
The method of claim 2, wherein the result providing unit
Providing profile information of the user who created the data when clicking on the data analysis result or when the cursor is positioned on the data analysis result
Big data analysis system based on human user profile analysis.
The method of claim 2, wherein the result providing unit
When the data analysis result is selected, providing the result of analyzing the profiles of users who generated the data
Big data analysis system based on human user profile analysis.
The method of claim 2, wherein the result providing unit
When a specific point in time is selected from the results analyzed over time, providing the result of analyzing the profiles of users who created data at the specific point in time
Big data analysis system based on human user profile analysis.
The method of claim 1, wherein the user profile analysis unit
Analyzing the profiles of the users using at least one of user information included in the data, user information inferred from the data, and user information provided by a service provider.
Big data analysis system based on human user profile analysis.
The method of claim 1, wherein the user profile analysis unit
When collecting data, analyzing the user's profile by collecting other data generated by the user who created the data
Big data analysis system based on human user profile analysis.
The method of claim 1, wherein the user profile analysis unit
When data is collected, the user's profile is analyzed using user information provided by the site where the data is disclosed.
Big data analysis system based on human user profile analysis.
The method of claim 1, wherein the data analysis unit
Limiting the range of the users and analyzing data generated by users included in the range based on the profiles of the users
Big data analysis system based on human user profile analysis.
Collecting online data generated by users;
For each of the users who created the collected data by extracting the profiles of the users using the data, and by additionally collecting other data created by each user, extracting user information through inference based on a plurality of data from the same author. Generating individual profile information;
Analyzing data generated by users meeting a specific search condition based on individual profile information of the users, and performing at least one of trend analysis and public opinion analysis on a limited group of users targeted for analysis; And
Providing a data analysis result for the data and providing group user profile information of users who generated the analyzed data
Big data analysis method based on user profile analysis comprising a.
The method of claim 10, wherein providing a result of analysis of the data and profile information of the users
Providing profile information of the user who created the data when selecting the analysis result for the data
Big data analysis method based on human user profile analysis.
The method of claim 10, wherein providing a result of analysis of the data and profile information of the users
Providing a result of analyzing the profiles of users who generated the data when selecting the analysis result for the data
Big data analysis method based on human user profile analysis.
The method of claim 10, wherein analyzing the data generated by the users based on the profiles of the users
Defining a range of the users based on the profiles of the users; And
Including the step of analyzing the data generated by users included in the range
Big data analysis method based on human user profile analysis.
The method of claim 10, wherein collecting the data generated by the users
When collecting data, including the step of collecting other data generated by the user who generated the data
Big data analysis method based on human user profile analysis.
The method of claim 10, wherein analyzing the profiles of the users
Analyzing the profiles of the users by analyzing information on the profile included in the data
Big data analysis method based on human user profile analysis.
The method of claim 10, wherein analyzing the profiles of the users
Checking data recognized as being created by the same user and analyzing the users' profiles based on the contents included in the data
Big data analysis method based on human user profile analysis.
The method of claim 10, wherein analyzing the profiles of the users
Analyzing the profiles of the users using user information provided by a service provider that manages the data
Big data analysis method based on human user profile analysis.

Images