KR102465711B1 - Dynamic prediction service system for congressional bill passing possibility using an analysis of a social data - Google Patents

Abstract

The present invention relates to a system for providing dynamic prediction services for the probability of passage of legislative bills in the National Assembly plenary session using social data analysis. The system includes: an open API bill information collection server which automatically collects information on the National Assembly bill from the National Assembly bill information system in real time; a National Assembly bill information database which stores the National Assembly bill information automatically collected in real time from the open API bill information collection server; and a prediction service providing server which includes a coding unit for automatically coding the National Assembly bill information stored in the National Assembly bill information database in real time, a probability of passage calculation unit for calculating the probability of each legislative bill passage in the National Assembly based on the result automatically coded by the coding unit, and a dynamic probability of passage prediction unit for dynamically predicting the probability of each National Assembly bill passage by analyzing environmental factors affecting the probability of passage of the bill submitted to the National Assembly in relation to the stored National Assembly bill information, wherein the environmental factors may include public opinion information identified based on social data collected in relation to the submitted bill.

Description

System for providing dynamic prediction service for the possibility of passing a bill in the plenary session of the National Assembly using social data analysis

The present invention relates to a service providing system for dynamically predicting the possibility of passing a bill in the plenary session of the National Assembly using social data analysis. In particular, the present invention dynamically calculates the passability of the bill in the plenary session of the National Assembly by analyzing environmental factors that affect the passability of the bill after the bill is submitted to the National Assembly, and visualizes the calculation result. It relates to a dynamic prediction service providing system that provides a serviceable platform and the possibility of passing a bill in the plenary session of the National Assembly using social data analysis.

The modern democratic government follows the principle of separation of powers to prevent the concentration of state power and guarantee the people's freedom and equality. As the influence of informal participants in the decision-making process of policy alternatives increases and the open and democratic policy-making method is strengthened due to the decentralization of the executive branch, the legislative branch has the right to enact laws, check the government, investigate the state affairs, and appoint and appoint important public officials. It has the right to consent and autonomous authority regarding internal matters of the National Assembly. The center of the legislative process is the review of the bill by the standing committee, and the stage of the plenary session is a legally important step in the final decision-making process of the bill. Its purpose is to legislate the political and policy decisions most in line with the will of the people.

At this time, a method of disclosing the data obtained by identifying and analyzing the political inclination of each member of the National Assembly and informing the degree of agreement on the bill at issue has been researched and developed. Published on November 02, 2015), a configuration is disclosed to analyze members of the National Assembly and political parties with similar voting tendencies to the user based on the voting results of the National Assembly and the results of the user's mock vote, and provide them to the user.

However, it is not possible to predict whether a bill that has passed the examination of the Standing Committee can be passed at the plenary session even if the above-described composition of the prior art is used. In terms of the organization of the parliament, we have continuously expanded the legislative support organization that provides legislative information to lawmakers so that it can develop into a policy-making parliament for productive legislative activities. It is a reality that it is far from coming up with a plan to keep the administration in check. Members of the National Assembly are constitutionally legislative and representative bodies of the people. However, the tension between the professional legislative function and the democratic representative function is high in that the elected members of the National Assembly are easily influenced by political ideology and public opinion rather than legislative expertise. Given that being exposed is common, it is becoming more difficult to predict the passability of this meeting. Therefore, it is required to research and develop a method of predicting the passage of a bill considering political ideology, party direction, and political situation.

On the other hand, as parliamentary democracy develops, the authority of the National Assembly is gradually strengthened, and the strengthening of the legislative authority of the National Assembly is particularly remarkable. As a result, the average number of bills submitted per day in the 14th National Assembly increased from 1.3 cases to an average of 27.4 cases per day in the 21st National Assembly. As the number of bill submissions soars, predicting the passage of bills is of great interest to all because the passage of bills has a great influence on the lives of companies, institutions, organizations, and individuals.

In this regard, conventionally, there is no technology that can predict the passability of a bill submitted to the National Assembly. As time elapses after submission, it is impossible to capture and predict the change in the possibility of passage according to changes in the overall environment surrounding the bill.

Since legislative bills are similar to living organisms, the possibility of a legislative bill that had a remarkably high possibility of passage depending on changes in the political and social environment surrounding the bill, that is, public opinion, can drop dramatically, and vice versa. Since it is difficult to predict the possibility of passage of a bill with such high variability, it is a reality that companies, institutions, and organizations that have an interest or influence with the bill in question respond by investing a lot of human and material resources. In addition, with the development of the Internet and SNS, the speed of formation and dissemination of opinions and public opinion on any issue is getting faster and changes occur frequently, making it more difficult to dynamically predict the possibility of passing a bill. Therefore, research and development of a method of dynamically predicting the passage of a bill by collecting and analyzing opinions or public opinion surrounding the bill is urgently required.

This application is intended to solve the above-mentioned problems of the prior art, considering that the passageability of the submitted bill may change according to changes in the overall environment surrounding the bill as time elapses after the submission of the bill, social data The purpose of this is to provide a system that dynamically predicts the passage possibility of the National Assembly plenary session of a bill using social data analysis, which can dynamically predict the passage possibility of a bill by collecting and analyzing opinions or public opinion surrounding the bill based on

This application is intended to solve the above-mentioned problems of the prior art, and derives a kind of parameter called emotional data represented by opinions on passing or not passing the bill by using data that automatically collects and analyzes a large amount of social data, including news. and predict the possibility of passing the bill in the plenary session of the National Assembly by conducting continuous learning and testing using an artificial intelligence algorithm. The purpose of the present invention is to provide a dynamic predictive service system for the passage of a bill in the plenary session of the National Assembly using social data analysis, which can predict the dynamic value of the passage possibility of a bill that changes from moment to moment.

However, the technical problem to be achieved by the embodiments of the present application is not limited to the technical problems described above, and other technical problems may exist.

As a technical means for achieving the above technical task, the system for providing a dynamic prediction service for the possibility of passing a bill in the plenary session of the National Assembly using social data analysis according to an embodiment of the present application automatically collects information on the National Assembly bill in real time from the National Assembly bill information system Open API bill information collection server; a National Assembly bill information database for storing the National Assembly bill information automatically collected in real time from the open API bill information collection server; And a coding unit for automatically coding the National Assembly bill information stored in the National Assembly bill information database in real time, a passability calculation unit for calculating the passage possibility of each National Assembly bill based on a result automatically coded in the coding unit, and the stored National Assembly. A prediction service providing server including a dynamic passability forecasting unit that dynamically predicts the possibility of passing the National Assembly of each bill by analyzing environmental factors affecting the passage of the bill submitted to the National Assembly in relation to bill information; , The environmental factor may include public opinion information identified based on social data collected in relation to the submitted bill.

In addition, the prediction service providing server may include a keyword extraction unit for extracting a keyword of the bill from the contents of the stored information on the National Assembly bill; a social data collection unit that collects social data related to the extracted keywords of the legislative bill from a social network service providing server and stores them in a social database; and a sentiment analysis unit deriving sentiment analysis data corresponding to the social data as the public opinion information by performing sentiment analysis on the social data stored in the social database.

In addition, the passability calculator may predict the passability of each legislative bill in the National Assembly through at least one type of artificial intelligence algorithm and machine learning.

In addition, the prediction service providing server selects predetermined information among the National Assembly bill information stored in the National Assembly bill information database as at least one factor influencing the possibility of passing the National Assembly plenary session, and passes the at least one factor and the passage. It may further include a modeling unit that builds a prediction model for passing the plenary session of the National Assembly by learning and testing the relationship between possibilities with at least one artificial intelligence algorithm and machine learning.

The above-described problem solving means are merely exemplary and should not be construed as intended to limit the present disclosure. In addition to the exemplary embodiments described above, additional embodiments may exist in the drawings and detailed description of the invention.

According to any one of the above-described problem solving means of the present invention, by providing a system for providing a dynamic prediction service for the passage of the National Assembly plenary session of a bill using social data analysis, as time elapses after the submission of the bill, the overall environment surrounding the bill Considering that the possibility of passing a proposed bill may change according to changes, the possibility of passing a bill can be dynamically predicted by collecting and analyzing opinions or public opinion surrounding the bill based on social data.

According to any one of the above-described problem solving means of the present invention, by providing a system for providing a dynamic prediction service for the passage of a bill in the plenary session of the National Assembly using social data analysis, data obtained by automatically collecting and analyzing a large amount of social data including news It derives a kind of parameter called sentiment data, which is represented by opinions on passing or not passing the bill, and predicts the possibility of passing the bill in the plenary session of the National Assembly by conducting continuous learning and testing using artificial intelligence algorithms, and this sentiment data parameter is a floating variable, not a fixed variable, which is continuously updated according to the results of social data collection and analysis, and can predict the dynamic change value of the legislative passability that changes from moment to moment.

According to any one of the above-described problem solving means of the present invention, by providing a service providing system for dynamically predicting the possibility of passing a bill in the plenary session of the National Assembly using social data analysis, specific information formed on social networks such as news, SNS, blogs, and communities. The possibility of passing a bill in the plenary session of the National Assembly is predicted by considering the parameters (parameters related to environmental factors) including the amount of social data (buzz) and trends of positive or negative opinions flowing on social data, etc., and the parameters are fixed. It is possible to dynamically predict the possibility of passing a legislative bill by continuously and automatically updating it with a floating variable rather than a variable, and through this, the public, including companies and institutions affected by the passage of the bill, can express their opinions on the legislative process more effectively at a lower cost. By enabling development and monitoring, it can contribute to the democratic and open operation of the National Assembly, the representative body of the people.

However, the effects obtainable herein are not limited to the effects described above, and other effects may exist.

1 is a diagram for explaining a system for providing a service for dynamically predicting the passability of a legislative bill in a plenary session of the National Assembly using social data analysis according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating a prediction service providing server included in the system of FIG. 1 .
3 and 4 are diagrams for explaining an embodiment in which a service for dynamically predicting the passage of a legislative bill in the plenary session of the National Assembly using social data analysis according to an embodiment of the present invention is implemented.
5 is an operational flow chart illustrating a method for providing a service for dynamically predicting the passability of a legislative bill in the plenary session of the National Assembly using social data analysis according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail so that those skilled in the art can easily practice the present invention with reference to the accompanying drawings. However, the present invention may be embodied in many different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only the case where it is "directly connected" but also the case where it is "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, this means that it may further include other components, not excluding other components, unless otherwise stated, and one or more other characteristics. However, it should be understood that it does not preclude the possibility of existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

As used throughout the specification, the terms "about", "substantially", etc., are used at or approximating that value when manufacturing and material tolerances inherent in the stated meaning are given, and do not convey an understanding of the present invention. Accurate or absolute figures are used to help prevent exploitation by unscrupulous infringers of the disclosed disclosure. The term "step of (doing)" or "step of" as used throughout the specification of the present invention does not mean "step for".

In this specification, a "unit" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Further, one unit may be realized using two or more hardware, and two or more units may be realized by one hardware. On the other hand, '~ unit' is not limited to software or hardware, and '~ unit' may be configured to be in an addressable storage medium or configured to reproduce one or more processors. Thus, as an example, '~unit' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. Functions provided within components and '~units' may be combined into smaller numbers of components and '~units' or further separated into additional components and '~units'. In addition, components and '~units' may be implemented to play one or more CPUs in a device or a secure multimedia card.

In this specification, some of the operations or functions described as being performed by a terminal, device, or device may be performed instead by a server connected to the terminal, device, or device. Likewise, some of the operations or functions described as being performed by the server may also be performed by a terminal, apparatus, or device connected to the server.

In this specification, some of the operations or functions described as mapping or matching with the terminal mean mapping or matching the terminal's unique number or personal identification information, which is the terminal's identifying data. can be interpreted as

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

1 is a diagram for explaining a system 1 for providing a service for dynamically predicting the passability of a legislative bill in the plenary session of the National Assembly using social data analysis according to an embodiment of the present invention. Hereinafter, for convenience of description, the system 1 for providing a service for dynamically predicting the passability of the National Assembly plenary session of a bill using social data analysis according to an embodiment of the present invention may be referred to as the present system 1. In addition, the prediction service providing server 300 included in the present system 1 is the prediction service providing server 300 according to an embodiment of the present invention, and hereinafter may be referred to as the present server 300 for convenience of explanation. .

Referring to FIG. 1, the system 1 for providing a dynamic prediction service for the possibility of passing a bill using social data analysis in the plenary session of the National Assembly includes at least one user terminal 100, a prediction service providing server 300, and at least one open API bill. It may include an information collection server 400 and at least one National Assembly bill information database 500 . However, since the dynamic prediction service providing system 1 of the possibility of passage of a bill using social data analysis in FIG. 1 is only one embodiment of the present invention, the present invention is not limitedly interpreted through FIG. 1 .

The present system 1 and this server 300 may provide users with a service for predicting the possibility of passing a bill submitted to the National Assembly in the plenary session of the National Assembly (ie, a service for predicting the possibility of a bill to pass in the plenary session of the National Assembly). In particular, the present system 1 and present server 300 can provide static/dynamic prediction services by performing static prediction and dynamic prediction in relation to the possibility of passing a bill in the plenary session of the National Assembly. Here, the static prediction may be made by calculating the passability of each legislative bill to the National Assembly based on the automatically coded result in the coding unit by the passability calculation unit in the present server 300 . On the other hand, the dynamic prediction may be made by dynamically predicting the passage possibility of each National Assembly bill by the dynamic passability prediction unit in the present server 300 by analyzing environmental factors based on social data. That is, the information calculated (predicted) by the passability calculation unit in this server 300 means static prediction information on the passageability of the bill in the National Assembly, and is calculated (predicted) by the dynamic passability prediction unit in this server 300. The information to be used may mean dynamic prediction information on the possibility of passage of a bill in the National Assembly.

In this application, the prediction service for the passage of the National Assembly plenary session of the bill provided by the system 1 and the server 300 refers to the static prediction service and social data analysis of the passage possibility of the National Assembly plenary session of the bill based on the coded result. It can be understood as a concept including a dynamic prediction service of the possibility of a bill passing the plenary session of the National Assembly, and accordingly, it can be referred to as a static/dynamic prediction service of the possibility of a bill passing the plenary session of the National Assembly.

Hereinafter, in describing the present invention, the system 1 is referred to as a system for providing a service for dynamically predicting the passage of a bill at the plenary session of the National Assembly using social data analysis, but is not limited thereto. Possibility of passing the plenary session of the National Assembly can be otherwise referred to as a static/dynamic predictive service providing system.

Each component of FIG. 1 is generally connected through a network (200). For example, as shown in FIG. 1 , at least one user terminal 100 may be connected to a prediction service providing server 300 through a network 200 . In addition, the prediction service providing server 300 includes at least one user terminal 100, at least one open API bill information collection server 400, and at least one National Assembly bill information database 500 through the network 200. can be connected In addition, at least one open API bill information collection server 400 may be connected to the prediction service providing server 300 through the network 200 . And, at least one National Assembly legislative bill information database 500 may be connected to the prediction service providing server 300 through the network 200 .

Here, the network refers to a connection structure capable of exchanging information between nodes such as a plurality of terminals and servers, and examples of such networks include a local area network (LAN) and a wide area network (WAN: Wide Area Network), the Internet (WWW: World Wide Web), wired and wireless data communications networks, telephone networks, and wired and wireless television communications networks. Examples of wireless data communication networks include 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), 5th Generation Partnership Project (5GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi , Internet (Internet), LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth (Bluetooth) network, NFC ( A Near-Field Communication (Near-Field Communication) network, a satellite broadcasting network, an analog broadcasting network, a Digital Multimedia Broadcasting (DMB) network, etc. are included, but not limited thereto.

In the following, the term at least one is defined as a term including singular and plural, and even if at least one term does not exist, each component may exist in singular or plural, and may mean singular or plural. It will be self-evident. In addition, the singular or plural number of each component may be changed according to embodiments.

At least one user terminal 100 passes at least one bill and at least one bill at the plenary session of the National Assembly using a web page, app page, program or application related to dynamic prediction service related to the possibility of passing a bill using social data analysis It may be a terminal that outputs possibilities. At this time, at least one user terminal 100 may output the probability of passing the plenary session of the National Assembly in % or output as visual data according to each % section. It may be yellow or green. However, it will be apparent that the visual data is not limited to the above-mentioned visual tools, and may include various visual data such as diagrams and graphs, and may be changed according to the embodiment.

Here, at least one user terminal 100 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, a laptop, a desktop, a laptop, and the like equipped with a navigation system and a web browser. In this case, at least one user terminal 100 may be implemented as a terminal capable of accessing a remote server or terminal through a network. At least one user terminal 100 is, for example, a wireless communication device that ensures portability and mobility, navigation, PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet ) may include all types of handheld-based wireless communication devices such as terminals, smartphones, smart pads, tablet PCs, and the like.

The prediction service providing server 300 may be a server that provides a dynamic prediction service web page, app page, program or application on the possibility of passing a bill using social data analysis in the plenary session of the National Assembly. In addition, the prediction service providing server 300 identifies and extracts at least one unique information included in the bill, and extracts a preset factor among the unique information, that is, a factor affecting the possibility of passing the National Assembly plenary session, It may be a server that predicts (dynamic prediction) the possibility of passing the plenary session of the National Assembly through the extracted factors. To this end, the prediction service providing server 300 predicts the passability of the plenary session of the National Assembly by learning and testing the relationship between at least one factor and passability with at least one artificial intelligence algorithm and machine learning to predict the passability of the plenary session of the National Assembly. It may be a server capable of conducting modeling to build a model. In addition, the prediction service providing server 300 is a server that codes to classify unique information within the legislative bills collected in real time or periodically so as to extract a predetermined factor among unique information, and identifies and extracts a predetermined factor among each unique information. can be

Here, the prediction service providing server 300 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, a laptop, a desktop, a laptop, and the like equipped with a navigation system and a web browser.

At least one open API bill information collection server 400 collects information from the National Assembly bill information system using a web page, app page, program, or application related to the dynamic prediction service of the possibility of passing the National Assembly plenary session of a bill using social data analysis can be a server At this time, the URL of the National Assembly Bill Information System is http://likms.assembly.go.kr/bill/main.do and provides open data as public information. Of course, it is possible to collect information from one open API bill. Here, at least one open API bill information collection server 400 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, a laptop, a desktop, a laptop, and the like equipped with a navigation system and a web browser.

At least one National Assembly legislative bill information database 500 is collected from the open API bill information collection server 400 using a web page, app page, program, or application related to a dynamic prediction service for the possibility of a bill passing the National Assembly plenary session using social data analysis. It can be a repository for storing information. Here, at least one National Assembly legislative bill information database 500 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, a laptop, a desktop, a laptop, and the like equipped with a navigation system and a web browser.

Figure 2 is a block configuration diagram for explaining the prediction service providing server 300 included in the present system 1 of Figure 1, Figures 3 and 4 are the present system 1 according to an embodiment of the present invention and It is a diagram for explaining an embodiment in which a dynamic prediction service for the passage of a bill in the plenary session of the National Assembly using social data analysis provided by the server 300 is implemented.

1 and 2, the prediction service providing server 300 (this server) includes a coding unit 310, a passability calculation unit 320, a visual representation unit 330, a modeling unit 340, and an error correction unit. (350). The passability calculator 320 may perform a static prediction on the passability of a bill in the plenary session of the National Assembly.

In addition, this server 300 is a keyword extraction unit 360, a social data collection unit 370, a social database 380, a sentiment analysis unit 390 and a dynamic passability prediction unit 391 may be further included. The dynamic passability prediction unit 391 may perform dynamic prediction on the passability of a bill in the plenary session of the National Assembly.

The prediction service providing server 300 according to an embodiment of the present invention or another server (not shown) operating in conjunction with at least one user terminal 100, at least one open API bill information collection server 400, and In the case of transmitting a dynamic prediction service application, program, app page, web page, etc. to at least one National Assembly legislative bill information database 500 on the possibility of passing a bill using social data analysis, at least one user terminal 100, at least One open API bill information collection server 400 and at least one National Assembly bill information database 500 dynamically predict the possibility of passing a bill in the National Assembly plenary session using social data analysis, a service application, program, app page, web page, etc. can be installed or opened. In addition, the service program is driven in at least one user terminal 100, at least one open API bill information collection server 400, and at least one National Assembly bill information database 500 using a script executed in a web browser. may be Here, the web browser is a program that allows users to use the web (WWW: World Wide Web) service, and means a program that receives and displays hypertext described in HTML (Hyper Text Mark-up Language). For example, Netscape , Explorer, Chrome, and the like. In addition, an application means an application on a terminal, and includes, for example, an app running on a mobile terminal (smart phone).

As mentioned above, this server 300 can perform static prediction and dynamic prediction in relation to the possibility of passing a bill in the plenary session of the National Assembly. Hereinafter, the static prediction technology will be described and then the dynamic prediction technology will be described. .

<Static prediction of the possibility of passing the bill in the plenary session of the National Assembly by this server 300>

Referring to FIG. 2, first of all, in order to predict (static prediction) the possibility of passing the plenary session of the National Assembly, it is necessary to look at what bills are proposed for the plenary session of the National Assembly. To this end, the open API bill information collection server 400 may automatically collect National Assembly bill information in real time from the National Assembly bill information system. As of the filing date, the url of the bill information system of the National Assembly is http://likms.assembly.go.kr/bill/main.do, and the bill type, processing status (processing/pending), name of the proposer or proposer , You can search the bill number, bill name, and number of proposals (how many National Assembly members there are), the list of recent plenary session bills and their contents (original text of bills), the list of recently received bills and their contents, and the agenda for plenary sessions. and its contents are disclosed respectively.

At this time, the collection is performed in real time using the open API bill information collection server 400. If you look at the list of recently received bills with reference to the site of the National Assembly Bill Information System with reference to FIG. . At this time, a system capable of predicting whether or not the list passes before the review by the plenary session of the National Assembly is required, and in one embodiment of the present invention, prediction is made based on this. The resolution list of this meeting, which has already resulted in the results, can be used as a training and test dataset for initial prediction modeling and data for re-learning the accuracy of prediction results. Real-time data is based on data prior to review by the plenary session of the National Assembly. At this time, since the review of the bill by the standing committee is premised as a step before the plenary session of the National Assembly, this process can also be predicted together. However, in the method according to an embodiment of the present invention, only the review by the plenary session of the National Assembly, that is, the final step will be used for convenience of description.

The open API legislation information collection server 400 may use an open API to collect the above-described data, among which a web crawler may be used. Here, a web crawler is a computer program that searches the World Wide Web in an organized and automated way. The work a web crawler does is called web crawling or spidering, and is a form of bot or software agent. There are two types of web crawlers: general web crawlers and distributed web crawlers. To explain the basic operation of a web crawler, it starts by getting a URL from the URL Frontier module and fetching a web page of that URL using the HTTP protocol. Then, the Fetch module stores the web page in temporary storage, the Parser module extracts text and links, and the text is sent to the Indexer. In the case of a link, whether it should be added to the URL frontier is determined by going through the Content Seen, URL Filter, and Duplication URL Element modules. At this time, since it is virtually impossible to crawl all web documents with a general web crawler, a distributed web crawler can be used more.

Distributed web crawlers are largely divided into two types, one of which is centralized and the other is P2P (or Fully-Distributed). A centralized distributed web crawler is a structure in which the URL manager acts as a server and the crawler acts as a client. After downloading the document from the crawler, extracting the OutLink URL and passing it to the URL manager, the URL manager checks whether it is the URL of the downloaded document and removes the URL duplication. In other words, the URL manager replaces the URL duplication and URL management in general web crawlers. On the other hand, in the P2P method, each crawler has a completely independent structure. In the P2P method, each crawler operates like a normal web crawler. Each crawler downloads documents, extracts OutLink URLs, and removes duplicate URLs. Each crawler operates independently. To do this, the list of downloaded URLs managed by each crawler must be mutually exclusive. Otherwise, the same document will be downloaded from different crawlers. As a way to solve this problem, each crawler can divide and manage URL domains to be downloaded exclusively from each other. That is, it manages only the download domain itself and passes the rest of the URLs to other crawlers. When using this method, each crawler can operate independently.

Next, web content must be extracted. Web content extraction technology provides a function to automatically extract the product name, author, publication date, text, and detailed information within the text of the resell product, which is the content to be used for information analysis, from the web document. . The web content extraction system is a device that extracts only content by automatically producing rules for extracting content. It includes a rule generator that automatically creates content extraction rules and a navigation content eliminator that removes navigation content from a given web document. , It can be composed of a content extractor (Core Context Extractor) that extracts content through content extraction rule keyword similarity comparison. Of course, it is also possible to collect real-time data in various ways other than the above-described crawling method, and is not limited to those listed and is not excluded for reasons not listed.

Next, the National Assembly bill information database 500 may store the National Assembly bill information automatically collected from the open API bill information collection server 400 in real time. At this time, the DBMS (DataBase Management System) must satisfy ACID, that is, Atomicity, Consistency, Isolation, and Durability, with minimal overhead under any circumstances. To this end, DBMS uses logs, but logging based on ARIES (Algorithms for Recovery and Isolation Exploiting Semantics) requires log buffer writes to be sequentially written, which is a lot of work to sequentially write log writes in a multi-core processor environment. wastes performance. To improve this, space for log writing was allocated in the central log buffer, and then worker threads copied logs to the allocated space in parallel. The space has been allocated, but you need to find the space where the log copy has not been completed.

MySQL 8.0 solves this problem so that worker threads record the LSN (Log Sequence Number) of the log that has been copied in units of log records, and the thread performing the log write operation sequentially tracks the corresponding information to find free space. However, this method causes a problem that one log writing thread cannot track the log growth rate by multiple worker threads. To this end, the National Assembly bill information database 500 according to an embodiment of the present invention applies an optimization algorithm to the MySQL 8.0 logging engine to use a delay phenomenon mitigation method for finding empty space in the log buffer.

An optimization algorithm according to an embodiment of the present invention may use a Border-Collie algorithm. The Border Collie thread sequentially searches LSNs recorded in units of threads to find a Recoverable Logging Boundary (RLB). This is a read-optimal algorithm because it reads information as many as the number of worker threads, rather than reading information generated for every number of log records created by worker threads. Also, in the Border Collie algorithm, an operation of finding a range of sequentially written logs can be executed without a wait. The operation of the Border Collie algorithm is as follows.

The Border Collie Algorithm defines RLB, which is a boundary indicating that there is no empty space up to that point. In addition, immediately before and after copying the log to the buffer, the LSN at the corresponding time is stored for each worker thread, and a state flag indicating whether the worker has started or finished copying to the buffer is also Save. The thread of the Border Collie Algorithm reads the latest LSN to obtain the RLB and designates it as a cutoff flag. Then, it traverses all user threads and calculates the RLB. At this time, if the status flag of the user thread indicates that the log is currently being written, the corresponding user thread may have created a log hole, so the smallest LSN of currently running workers is selected as the next RLB candidate. do.

However, if the database worker's status flag reports that the current worker is not writing logs, it means that this worker did not create a log hole until the cutoff flag. In this case, the cutoff flag is selected as the RLB candidate. Then, by comparing each candidate, a smaller value is selected as the RLB. Also, to solve transaction processing delay in MySQL 8.0, Latency Hiding is applied. In the Latency Hiding technique, after a worker thread copies a log to a buffer, it immediately performs an operation for the next transaction. Reporting on commits is executed by calling the callback function after the log is flushed.

When the border collie algorithm and the delay hiding technique are applied to the MySQL 8.0 logging engine, in MySQL 8.0, the Log Writer traverses the buffer where the LSN information is posted, looking for the range of serially written logs, Logs in the corresponding range are saved to disk. In one embodiment of the present invention, in order to apply the Border Collie algorithm to the logging engine of MySQL 8.0, a module that searches logs stored in the buffer by sequentially tracing the global buffer where LSN information is posted finds the LSN corresponding to the RLB. Modules can be replaced. Specifically, a global buffer with as many slots as the number of workers is allocated, each worker thread is allocated this slot, and the worker threads update the LSN and status flag in the global buffer immediately before and after writing a log to the log buffer. The thread traverses this slot, obtains the LSN corresponding to the RLB, and sets the log to be flushed up to this LSN.

Also, MySQL 8.0 does not start processing the next transaction until all logs of the current transaction are flushed. In order to apply the delay hiding technique to the MySQL 8.0 engine, a callback method can be defined that sends a commit completion message to THD (short for Thread), which is a class that is responsible for connecting with clients, and is currently working before performing the operation of the next transaction. The THD pointer and LSN of the in transaction can be pushed onto the queue. Then, after dequeueing the queue nodes with LSNs smaller than the currently flushed LSN, the callback function is called using the THD pointer in the deleted node.

In this way, when the border collie algorithm and the delay hiding technique are applied to the logging engine of the MySQL 8.0 environment of the DBMS, the amount of operations per second of database construction can increase. The reason why the performance difference increases as the number of working threads increases is that the log writer thread's log hole search speed lags behind the log creation speed of multiple worker threads. Also, the Border Collie algorithm can flush logs more quickly by tracking log holes with a small delay from logs created by a large number of workers. In addition, if the border collie algorithm is applied during the workload of writing and reading, processing of the log generated by the write operation becomes faster, and the application of the delay hiding technique is followed by copying to the central log buffer until the log flush is completed. Instead of waiting for the operation of a transaction to be executed, it makes it possible to perform the operation for the next transaction, which allows more transactions to be processed at the same time. Accordingly, it is possible to write and read a plurality of legislative bills uploaded in real time at a high speed, and the delay in analysis results accordingly can be reduced.

Once the collection and storage have been completed, it is now necessary to analyze the bill and predict whether the bill can be passed in the plenary session of the National Assembly. First, the coding unit 310 of the prediction service providing server 300 may automatically code the National Assembly bill information stored in the National Assembly bill information database 500 in real time. At this time, the legislative bill is provided in hwp or pdf format with a certain format, and only the file format is different for both hwp and pdf, but the internal format is constant. In Figure 4a, among the recently proposed (not yet reviewed) bills, "Partial Amendment to the Basic Act on the Settlement of the Past for Truth and Reconciliation" is examined with reference to Figure 4b. , the main content of alternatives, bills, and new structure comparison tables are sequentially listed, which is unique information of bills. At this time, since each table of contents or format is unified, it is not difficult to extract each information, but coding work is required to label or tag unique information for automation, identify it, and extract information mapped to each unique information. . In addition, since not all unique information is used for pass prediction, it is necessary to select which kind or item of unique information to use among the unique information, which requires human intervention and manual selection, tagging, or identification. It should be done. Once set, continuous human intervention is not required because factors that affect unique information can be automatically extracted thereafter. However, it is obvious that the influencing factors are not fixed and can be flexible factors that can be changed according to the research results of all walks of life.

For example, at least one factor may include a factor according to the individual characteristics of the proposer and a factor according to the characteristics of the bill. In the case of the former, in general, in the legislative activities of members of the National Assembly, the seniority, party, form of election, and ideology of individual members are important variables that determine legislative productivity. These discussions focus on the individual characteristics and institutional characteristics of individual lawmakers, and the higher the seniority, the ruling party, and proportional representation, the higher the bill passing rate. On the other hand, the press and media, which developed as the grip of authoritarian institutions weakened after democratization, form a consensus with the public as a major actor indirectly influencing the policy-making process through power monitoring, information and education functions, and agenda setting functions. The main role and responsibility of the press and the media is to monitor, cover, and report on the legislative activities of elected officials. The following characteristics appear in the way the press and media reports reflect the political reality. Members who are most active in Congress will be reported more, more members with high news value will appear, and since the news media itself has a bias, more members who contribute to the interests of the media or fit well with the editorial direction will be exposed more. becomes In other words, the social influence of the press and the media is an important means to provide members of the National Assembly with an opportunity for re-election through raising awareness, and to provide the public with information on legislative activities, evaluation indicators, and timely information on policies. do.

The latter case is a factor in the nature of the legislation. Factors that affect the efficiency and productivity of legislation include individual and institutional factors as well as policy characteristics such as the content and nature of the policy. In the modern administrative state, due to the complexity, expertise, and diversity of policies, political conflicts between actors in the legislative process inevitably appear. In order for the legislative activities of the legislature to develop and function smoothly in a reasonable and efficient way, thorough preparation and planning, such as discussion and characteristics of the law, are required. The stages of public policy are divided into distributive, regulatory, and re-distributive types, and different political structures and processes appear for each type of policy. The objective of the policy affects the policy process. Crazy. Distributive policy refers to a policy in which the government provides services or goods to specific individuals or organizations, and includes government subsidies and subsidies. Regulatory policy is a policy that applies control to protect a large number or a specific group, and it is classified into competitive protection policy and protective regulatory policy. Competitive protection policy is a mixture of regulatory policy and distribution policy, and among many competitors, a specific group or individual is given the right to a specific service or a matter on which many interests are staked, and those involved try to gain the advantage through competition. A protective regulatory policy is intended to protect the majority of the population by imposing restrictions on individuals, businesses, etc. Redistribution policy is a policy that transfers wealth or power from a group that owns a lot to a group that does not, through income redistribution, which can cause conflict between classes.

Factors that can be extracted from the former and the latter can be summarized as shown in Table 1 below, but are not limited thereto. Also, the factors are not limited to those listed and are not excluded for reasons not listed.

[Table 1]

If the factors that affect this are selected, these factors are searched and found within the bill, and +/- is performed according to the factors, the possibility of approval can be derived. Based on the automatically coded results, it is possible to calculate the possibility of passing each National Assembly bill. The passability calculation unit 320 may predict the passability of each National Assembly bill through at least one type of artificial intelligence algorithm and machine learning. At this time, at least one type of artificial intelligence algorithm and machine learning includes a Neural Network, a Hidden Markov Model (HMM), and a Support Vector Machine (SVM), but is not limited to the listed ones and is excluded for reasons not listed. For example, the method of predicting results using the above-mentioned factors as independent variables is to measure the seniority of a member as a personal characteristic factor i) by the number of elections up to the 21st National Assembly, and ii ) Next, if the party belonging to the representative initiating member is the ruling party, it is measured as 1, and the remaining members of the opposition party are measured as 0. In the case of lawmakers whose political party was changed in the middle, they can be classified as the party they belonged to the longest based on their term of office of 4 years. Third, iii) As for the form of election of members, refer to the National Assembly Handbook and the Internet National Assembly website for the year, and measure 0 in the case of local constituency and 1 in case of proportional representation, respectively. For reference, it is measured by the number of co-sponsoring members who participated in bills submitted to the standing committee. v) In the case of a member of the National Assembly who was a member of the judicial profession, which indicates the professional experience of the lawmaker, it is measured as 1 for a member of the National Assembly who has a career in the legal profession, and 0 for a member who does not. vi) It is measured as 1 if the former position or career of the member has professional connection with the relevant standing committee, and as 0 if the member does not. vii) As for the variables related to the frequency of press and media exposure of lawmakers, the frequency of press exposure of lawmakers is identified through at least one portal site. There are negative and positive contents in the report, but media exposure raises awareness of the public. , we can measure the frequency of reporting through the number of articles in which the member of the National Assembly appears. In order to exclude exposure of persons with the same name other than members of the National Assembly, the number of articles in which the name of the member is entered and the word member or National Assembly appears within 5 words contiguity in the title or text, but is not limited to this, can be measured. .

In the case of the latter i) In order to measure the issue intensity of the bill, we searched for bills initiated by members of the National Assembly during the 21st National Assembly posted on the portal site and collected the frequency of articles published in news, media, and all daily newspapers, and defined operationally. can drop ii) Budget-accompanying bills are measured as 1 as they are accompanied by a budget when the cost estimate is submitted, and 0 when they are not accompanied by a budget. Whether or not to submit a cost estimate due to budget requirements can be collected from the National Assembly Bill Information System. iii) For the 21 controversial bills, the name of each committee, the controversial bill, conflict, etc. are entered together as search terms, and the bills evaluated as controversial bills are evaluated as 1, and non-controversial bills are measured as 0.

In addition, it is deemed to have been passed in consideration of the fact that the original bill, the amended bill, and the bill to abolish the reflection of alternatives are being passed through the system of the bill information system of the National Assembly. Non-reflection of the law is classified as rejection, withdrawal, withdrawal, and rejection. In this study, which utilizes logistic analysis, the passage of the original draft, the revision, and the rejection of alternatives are considered as passed and measured as 1, and other laws not reflected as 0. In this way, in the case of selecting and extracting the influencing factors (factors) among the unique information and conducting learning on the cause (independent variable) and result (dependent variable) of each factor, the correlation can be derived, and which factor is the National Assembly It is possible to check whether it is a factor that affects the passage of the plenary session, and even when a bill is entered through a query, the factor that affects it is extracted, and the model in which the correlation with the dependent variable is learned In the case of using , the result can be predicted by the result value, and the passing rate can be quantified. Of course, it may vary depending on the embodiment, and it will be obvious that the above-described embodiment is merely an embodiment.

At this time, the passability (static/dynamic passability) of the National Assembly bill may be output in the range of 0 to 100, and the visual representation unit 330 may pass the National Assembly bill output from the passability calculation unit 320 to the National Assembly. When the probability is received as a number from 0 to 100, the possibility of passing the National Assembly bill can be converted into a three-color traffic light system mapped for each preset section and output. The user terminal 100 may access the prediction service providing server 300 to output information and display a calculation result of the possibility of passing the National Assembly of each bill of the National Assembly, and the user terminal may determine the possibility of passing the National Assembly of the bill of the National Assembly in advance. It can be converted into a three-color traffic light system mapped by section and output.

Of course, it is also possible to output the percentage or number along with the traffic light system. At this time, if the probability of passing the National Assembly falls within the range of 0 to 39, red indicates low impassability, if it falls within the range of 40 to 69, yellow indicates normal passability, and if it falls within the range of 70 to 100, it is colored red. It can be output in green, which indicates high, and users can see it in an intuitive color, so individuals, businesses, and organizations affected by changes in policies or laws can set up management strategies accordingly. For example, if the number of full-time employees is decided to be n as a requirement for corporate tax reduction, an employment plan can be established, and assuming that the condition for obtaining a national project is more than 100 million won in employment investment for 3 years, a budget of 100 million won You can prepare to secure it.

The modeling unit 340 selects predetermined information among the National Assembly bill information stored in the National Assembly bill information database 500 as at least one factor affecting the passage possibility of the National Assembly plenary session, and selects the at least one factor and the passage possibility. It is possible to build a predictive model for passing the plenary session of the National Assembly by learning and testing the relationship between at least one artificial intelligence algorithm and machine learning. In this case, as described above, it includes a Neural Network, a Hidden Markov Model (HMM), and a Support Vector Machine (SVM), but is not limited to the listed ones and is not excluded for reasons not listed.

At this time, case-based reasoning among data mining techniques is easy and simple to apply, and the update of the model can be performed in real time. Case-Based Reasoning (CBR) solves problems through cases most similar to new cases, so even if a new bill is entered after learning cases such as the independent variable and dependent variable described above, it is effectively reflected in the new bill. You can do it. In general, case-based reasoning may have relatively low predictive performance compared to other artificial intelligence techniques, but it can improve predictive performance by differentiating weights according to the importance of input variables. If the same weight is applied to all variables, important variables cannot be reflected more in the prediction model, and unimportant variables can be considered as important variables. can be reflected more in the model to improve the prediction performance.

In case-based reasoning, in order to select weights for input variables, opinions of expert groups may be reflected or weights may be applied through genetic algorithms. A model can be built, but the coefficients of the logit model can be applied to calculate the importance of variables and applied as weights for input variables. Since the coefficient of a variable in a logit model indicates how much it affects the dependent variable, it can be used as a weight for an input variable (independent variable) because the importance of the variable can be judged. Comparison of case-based reasoning (Pure CBR) model, in which the same weight is applied to input variables, and weighted CBR model, in which different weights are applied according to the importance of input variables, to improve the performance of the prediction model to which case-based reasoning is applied. And the performance of the predictive model, that is, the one with high accuracy can be used.

At this time, case-based reasoning predicts the result value of a new case by referring to cases applied in the past and their results, and solves the problem through specific knowledge of the extracted cases by partially extracting past cases that are most similar to the new cases. way to do it Case-based reasoning can be applied in fields with relatively little data and can be easily applied in complex or less structured fields. In case-based reasoning, even if a new case is added to the database, the model can be updated immediately without going through a special learning process. In general, case-based reasoning consists of five steps. i) If a new case is given as an input case, the case base is searched and the previous case most similar to the new case is retrieved. ii) Referring to the results of the extracted cases, iii) the recommendation results for the new cases are adapted (adaptation), and iv) the recommendation results are verified (validation). v) The case base is updated so that the new results presented through the verification step are saved as new cases and reused for future problem solving.

Similarity is measured to extract cases similar to the new input case from the case base, and a Euclidean distance (D) can be used, which can be expressed as in Equation 1 below. The k-nearest neighbor method can be used as a method of finding the existing case closest to the input case by the similarity criterion.

[Equation 1]

In the case-based reasoning technique, it is important how to assign weights to each attribute along with the attribute selection to measure similarity. If attributes with little relevance to the target variable are used with the same importance as attributes with high relevance, they can have a negative impact on the prediction performance. It can improve the predictive performance of grounded inference models. The weighted Euclidean distance according to the importance of each variable is calculated as shown in Equation 2 below.

[Equation 2]

Based on this, the modeling unit 340 may apply case-based reasoning to build a prediction model for passing the plenary session of the National Assembly. In order to improve the performance of the National Assembly's plenary passability prediction model based on case-based reasoning, a logit model with high usability is used to determine the weight of the case-based reasoning model. Therefore, two models of case-based reasoning are used to improve prediction performance by applying low weights to input variables with little relevance and high weights to input variables with high relevance. One is a case-based reasoning (Pure CBR) model that builds a predictive model, and the other is a case-based reasoning (Weighted CBR) model that builds a predictive model by applying different weights according to the importance of input variables. In addition, the prediction performance of the weighted CBR and the artificial neural network model and the SVM model can be compared and the model with the best performance can be used.

Logit's Stepwise method can be used to select the input variables of the predictive model, and the input variables (independent variables) are selected with the possibility of the bill passing the National Assembly plenary session as the dependent variable, and the selected input variable To apply the weight, the coefficient of the logit can be applied to calculate the weight for each variable. Since the logit coefficient indicates the degree to which each variable affects the dependent variable, the relationship between the input variable and the dependent variable can be identified, and a weighted CBR model can be constructed by applying these as weights for the input variables of case-based inference. have. The weight can be calculated by dividing the sum of the logit coefficient and importance of all independent variables by the logit coefficient and importance of each independent variable as shown in Equation 3 below.

[Equation 3]

Here, xi means the i-th logit coefficient and the importance of the independent variable. The independent variable selected from the logit has a negative effect on the dependent variable, for example, the more the bill is accompanied by the budget (independent variable), the lower the probability of passing the bill (dependent variable), so it has a negative effect, so the logit coefficient The weight can be calculated using the absolute value of Weights are applied to each independent variable to calculate the Euclidean distance between independent variables, similar cases are extracted through the weighted Euclidean distance, and case-based inference can be applied using the k-nearest neighbor method. By differentiating the weights for important variables, significant similar cases are found even in an imbalanced data environment, thereby improving the performance of the predictive model and at the same time enabling a robust model according to the imbalance ratio.

The error correction unit 350 extracts at least one factor from the National Assembly legislative bill information that is automatically collected in real time, inputs it as a query, and predicts the output through the input query when driving the predictability model for the passage of the plenary session of the National Assembly. By checking the data and whether or not it actually passed, the error part and the cause of the error can be identified. For example, if the probability of passage of the Animal Protection Act was 89%, but it did not pass, it is necessary to find the cause and the causative pattern and conduct error correction. At this time, the pattern that causes the unintelligence is found by human intervention, and when the cause of the error and the pattern are labeled, it becomes possible to learn with a learning algorithm. If these causes and patterns are found and causal relationships, that is, independent variables and dependent variables are selected and learned based on case reasoning, the recurrence rate of errors can be lowered.

Hereinafter, an operation process according to the configuration of the above-described prediction service providing server of FIG. 2 will be described in detail with reference to FIGS. 3 and 4 as examples. However, it will be apparent that the embodiment is only any one of various embodiments of the present invention, and is not limited thereto.

Referring to Figure 3a, (a) the prediction service providing server 300 collects information from the open API bill information collection server 400 and stores it in the National Assembly bill information database 500, (b) unique information from each bill Recognizes, extracts factors affecting passability among unique information, (c) calculates passability as an output value after using it as an input (Query) of the passability prediction model of the National Assembly plenary session, (d) user terminal 100 In order to intuitively understand the legislative bill and its passability, it can be displayed in red (R), yellow (Yello, Y), and green (G), respectively.

To this end, the prediction service providing server 300 needs to model the passability prediction model of the plenary session of the National Assembly. First, by using the bill for which the result has already been derived, the rejection and approval information are classified, and (b) the passability among the unique information After learning the causal relationship by manually or automatically extracting the factors influencing the independent variable and the dependent variable by dividing it into an independent variable and a dependent variable, (c) when learning is completed with the learning dataset, the experiment is conducted with the test dataset and modeling with an algorithm with high predictive rate. through the modeling process. In addition, (d) when an error occurs in the predicted result through real-time prediction, re-learning can be performed so that the error does not occur by inputting the cause and pattern of the error.

Matters that have not been described in the method for providing a dynamic prediction service for the possibility of passing a bill in the plenary session of the National Assembly using social data analysis of FIGS. Since the content described for the providing method is the same as or can be easily inferred from the description, the following description will be omitted.

As described above, the present system 1 and this server 300, the influence through the frequency of exposure to the press and media of individual members, the intensity of issues of the bill through the frequency of exposure to the press and media of the bill, and the personal characteristics of the representative proposer The possibility of passing the bill in the plenary session of the National Assembly is predicted (statically predicted) by considering parameters including seniority, election type, number of co-sponsors, issue intensity of the bill, which is a characteristic factor of the bill, and whether or not it is accompanied by a budget. Parameters are constantly updated as floating variables rather than fixed variables, so that at least one type of artificial intelligence algorithm is used to continuously learn and test to increase the prediction accuracy, and information on the National Assembly bill is provided in real time from the National Assembly bill information system. By automatically collecting and checking the approval in real time, the error rate can be reduced by comparative learning between predicted results and actual results. This system 1 and this server 300 can contribute to the democratic and efficient operation of the National Assembly, a representative institution of the people.

Hereinafter, the dynamic prediction technology of the possibility of passing a bill in the plenary session of the National Assembly will be described in more detail.

<Dynamic prediction of the possibility of passing a bill in the plenary session of the National Assembly by this server 300>

As described above, the present server 300 may include a coding unit 310, a passability calculation unit 320, a visual representation unit 330, a modeling unit 340, and an error correction unit 350. In addition, this server 300 includes a keyword extraction unit 360, a social data collection unit 370, a social database 380, and a sentiment analysis unit 390 to perform dynamic prediction on the possibility of passing a bill in the plenary session of the National Assembly. ) and a dynamic passability prediction unit 391 may be further included.

Based on the result automatically coded by the coding unit 310, the passability calculation unit 320 may calculate the passage possibility of each National Assembly bill at the time when each National Assembly bill is submitted to the National Assembly, and at this time, the calculated National Assembly bill Passability may be a static predictive value (ie, passability static predictive value).

The dynamic passability prediction unit 391 analyzes environmental factors affecting the passability of the bill submitted to the National Assembly as time elapses after the submission of the bill to the National Assembly, and the National Assembly stored in the National Assembly bill information database 500. The possibility of passing the National Assembly of each National Assembly bill corresponding to the bill information can be dynamically predicted.

Here, the submitted bill refers to a bill corresponding to the bill information of the National Assembly automatically collected from the open API bill information collection server 400 in real time and stored in the National Assembly bill information database 500 as the bill is submitted to the National Assembly. can

In addition, the environmental factor may include public opinion information (change information of public opinion, change trend information of public opinion) identified based on social data collected in relation to the submitted bill.

Specifically, the keyword extraction unit 360 may extract a keyword from the contents (main content) of the National Assembly bill information stored in the National Assembly bill information database 500 . The keyword extraction unit 360 may extract a keyword of a legislative bill for each of a plurality of pieces of information on a plurality of legislative bills stored in the National Assembly bill information database 500 .

The legislative keyword refers to a keyword indicating which subject the bill (legal bill information of the National Assembly) is related to, and may be referred to as a representative keyword or a subject keyword representing the bill. Keyword extraction unit 360, for example, various data analysis techniques (eg, semantic analysis, keyword-specific By applying frequency analysis, etc.), it is possible to extract legislative keywords (representative keywords) corresponding to the National Assembly bill information. The keyword extraction unit 360 may extract at least one keyword for each legislative bill information of the National Assembly.

For example, if the main content of the information on the bill of the 1st National Assembly among the information on the bills of the National Assembly is related to the 'partially amended bill of the Act on the Promotion and Foundation of Drone Utilization', the keyword of the bill corresponding to the information on the bill of the 1st National Assembly is For example, it may be the same as 'drone'. As another example, if the main content of the bill information of the 2nd National Assembly is related to the 'Partially Amended Act on the Installation and Use of Sports Facilities', the keywords for the bill corresponding to the bill information of the 2nd National Assembly include, for example, 'sports facilities', ' Installation of sports facilities' and 'installation of golf course' may be included.

The social data collection unit 370 collects social data (social data related to the legal proposal keywords) related to the keywords of the legislation extracted by the keyword extraction unit 360 from a social network service providing server (not shown), and stores them in the social database 380. can be saved The social data collection unit 370 may automatically collect, in real time, at least one piece of social data related to the legislative keyword extracted from each legislative bill information of the National Assembly, and store the collected social data in the social database 380 for each of the plurality of pieces of information on the legislative bills of the National Assembly. Social data can be stored by matching it with the unique number of the relevant National Assembly bill information.

Here, the social network service providing server (not shown) may refer to a server of at least one portal site that enables sharing of information among users by uploading information (data) such as media and contents. For example, the social network service providing server (not shown) may include servers such as Naver, Google, Daum, Facebook, Twitter, and Cyworld, but is not limited thereto. Social data refers to various types of information such as media and content that exist on the social network service providing server, and may include, for example, data such as news, SNS (SNS posts), blogs, and communities (community posts). It is not limited to this.

Illustratively, let's assume that the keyword extraction unit 360 extracts three legislative keywords including 'sports facility', 'installation of sports facility', and 'installation of golf course' as the keywords of the legislation corresponding to the bill information of the 2nd National Assembly. . In this case, the social data collection unit 370 collects a plurality of social data related to 'sports facility', 'installation of sports facility', and 'installation of golf course' from a social network service providing server (not shown) and provides 2 It can be stored and managed in the social database 380 by matching with the unique number of the National Assembly bill information. At this time, the social data collecting unit 270 may collect all kinds of social data including at least one keyword of the legislative bill among the three keywords of the legislative bill and store them in the social database 380 .

The sentiment analysis unit 390 may derive sentiment analysis data corresponding to the social data as public opinion information (parameters of environmental factors) by performing sentiment analysis on the social data stored in the social database 380 . The sentiment analysis unit 390 may perform sentiment analysis on each of a plurality of social data for each legislative bill of the National Assembly stored in the social database 380, and thus derive sentiment analysis data for each social data.

The sentiment analysis unit 390 may perform sentiment analysis based on text information of social data (which may mean any one social data among a plurality of social data) to derive sentiment analysis data corresponding to the social data. have. Here, sentiment analysis data is the result of sentiment analysis on opinions corresponding to text information of social data, which is a positive opinion that is a passing opinion on a bill corresponding to social data (information on a bill in the National Assembly) (i.e., on social data It can be divided into positive opinions that the proposed bill will pass the National Assembly) or negative opinions that do not pass the bill (ie, negative opinions that believe that the submitted bill on social data will not pass the National Assembly). have.

Sentiment analysis data derived by the sentiment analysis unit 390 (that is, sentiment analysis data corresponding to social data) is information classified as passing opinion or non-passing opinion, so a bill related to social data is passed to the National Assembly. As time elapses after submission, it can be used as public opinion information that is formed (generated) in relation to the bill submitted to the National Assembly. That is, the sentiment analysis data derived by the sentiment analysis unit 390 is a parameter (public opinion information) of environmental factors that affect the passageability of a bill submitted to the National Assembly as time elapses after the bill is submitted to the National Assembly. can be used (considered) as

The sentiment analysis unit 390 derives sentiment analysis data in real time for each of a plurality of social data collected in real time in relation to the extracted legislative keyword, and the bill submitted based on the derived sentiment analysis data (legal bill related to social data) ) It may be possible to determine (analyze) information on changes in public opinion (change trend information in public opinion) surrounding it.

The sentiment analysis unit 390 may derive sentiment analysis data corresponding to the social data by performing sentiment analysis based on text information of the social data.

For example, if the social data is a news article, the sentiment analysis unit 390 recognizes text information in novel data (for example, a news article), and recognizes text information (which may include not only Korean but also foreign languages). After separating the multilingual text) by language, sentiment analysis is performed, and a multilingual sentiment analysis process of synthesizing polarity values may be performed. To this end, the emotion analysis unit 390 divides and stores Japanese, Chinese, Korean, and English words, that is, words, in units through a process of segmenting sentences for each input language. For example, Rhino 1.0, a Korean morpheme analyzer, can be used, Jieba can be used as a Chinese segmentation tool, and Chinese Stanford Tagger among Stanford POS Taggers can be used as a Chinese morpheme analyzer.

Next, the sentiment analysis unit 390 returns the words of the four countries to the morpheme analyzer and counts the number of nouns, verbs, adjectives, and adverbs included in the input sentences of the four countries. In addition, the emotion scores of the four languages stored in word units are obtained through the four language emotion dictionaries. In addition, since emoticons that frequently appear in comments or comments have emotions, they can be added to multilingual sentiment analysis. The emoticon sentiment dictionary can use Emoji Sentiment Ranking 1.0. Finally, the emotional scores of all multilingual languages are obtained through the emotional scores of four languages and emoticons, and the number of nouns, verbs, adjectives, adverbs, roots, and emotional emoticons.

A specific process of the multilingual sentiment analysis process may be as follows, for example. The technique for multilingual sentiment analysis can be largely composed of three steps. First, when the sentiment analysis unit 390 receives a sentence for sentiment analysis, it divides the sentence into word units, determines the morpheme of each word through a morpheme analyzer, and counts the number of nouns, verbs, adjectives, and adverbs among the morphemes. . In particular, the sentiment analysis unit 390 adds and counts root words in addition to nouns, verbs, adjectives, and adverbs because the word composition of Korean is different from that of four languages including Chinese. In addition, the emotion analysis unit 390 counts the number of emoticons when the emotion value of the emoticon is positive or negative. At this time, taking Korean and Chinese as an example, Korean can be expressed as ΣWk, Chinese as ΣWc, and emoticon as ΣEc.

In the second step, the emotion analysis unit 390 may perform emotion analysis through the emotion dictionary. If the separated word is a word belonging to the sentiment dictionary, the sentiment value of the word is reflected as a sentiment score in the sentiment dictionary. This can be expressed as Equation 4 and Equation 5. Sj is the sentiment score of a specific word, Wj is a specific word obtained through the segmentation process of phrases composed of multiple languages, and SDW is a sentiment word included in the sentiment dictionary. In particular, Equation 5 satisfies the condition that a specific word is not included in a word in the emotional dictionary. In this case, the sentiment score Sj of the specific word becomes 0.

[Equation 4]

[Equation 5]

The emotion analysis unit 390 finally calculates the total emotion scores obtained for each language as nouns, verbs, adjectives, adverbs, roots (only for Korean), and the number of emoticons with positive and negative sentiment values (only for emoticons). After dividing by the sum of , the multilingual emotion score to be finally obtained is obtained by reflecting the result obtained and the weight assigned to each language. This can be calculated through Equation 6. Wc is the weight occupied by Chinese, Wk is the weight occupied by Korean, and We is the weight occupied by emoticons. Sc is the sum of all Chinese sentiment scores Sj, Sk is the sum of all Korean sentiment scores Sj, as in Chinese, and Se is the sum of all emoticon sentiment scores Sj. S is the multilingual emotional score.

[Equation 6]

That is, the emotion analysis unit 390 may derive emotion analysis data (sentiment score) corresponding to social data by finally deriving S from Equation 6 through the above-described process. According to the derived sentiment analysis data S, the sentiment analysis unit 390 derives (extracts) whether the opinion corresponding to the text in the social data is courtly emotion (ie, passing opinion) or negative emotion (ie, non-passing opinion). ) can be done. The sentiment analysis unit 390 performs sentiment analysis on a plurality of social data acquired in real time, and can grasp change trend information of public opinion from the result of the sentiment analysis.

According to this, the sentiment analysis unit 390 relates to information on the second National Assembly bill information as an example among the plurality of pieces of information on the bill of the National Assembly stored in the bill information database 500 of the National Assembly. , 'sports facility', 'installation of sports facility', 'installation of golf course'), sentiment analysis can be performed on each of a plurality of social data collected in real time related to each. That is, the sentiment analysis unit 390 each of a plurality of social data collected in relation to 'sports facility', which is the keyword of the first legislation, a plurality of social data collected in relation to 'installation of sports facility', which is the keyword of the second legislation, respectively; Sentiment analysis data (positive or negative result data) for each social data can be derived by performing sentiment analysis on each of the plurality of social data collected in relation to the third legislative keyword, 'golf field installation'. have.

Thereafter, the sentiment analysis unit 390 is based on the sentiment analysis data derived for each of the plurality of social data collected in relation to the bill information of the second National Assembly, and quantified public opinion information corresponding to the bill information of the second National Assembly (ie, public opinion change information, public opinion change trend information) can be derived. For example, the quantified public opinion information may be information expressed as a percentage of each positive opinion and negative opinion among the total number of a plurality of social data, or information expressed as percentage information of positive opinions (or negative opinions) However, it is not limited thereto, and any expression method of information capable of expressing the result of a comprehensive opinion of sentiment analysis data derived for each of a plurality of social data can be applied. As an example, if 100 pieces of social data are collected in relation to the bill information of the 2nd National Assembly, 70% of which are positive opinions and 30% are negative opinions, the quantified public opinion information is '7:3', '70:30' ', 'positive 70%', 'negative 30%', etc.

When public opinion information (in particular, quantified public opinion information) corresponding to each of the plurality of National Assembly bill information is derived by the sentiment analysis unit 390, then the dynamic passability prediction unit 391 is derived by the sentiment analysis unit 390. Public opinion information by information on the National Assembly bill derived by the sentiment analysis unit 390 and the possibility of passing the National Assembly of each National Assembly bill calculated by the passability calculation unit 320 (static Based on the predicted passability value), it is possible to derive a dynamic passability prediction value that dynamically predicts the passageability of each National Assembly bill (ie, the result value of the dynamic passageability prediction value of each National Assembly bill).

At this time, the dynamic passability prediction unit 391 includes public opinion information for each National Assembly bill information derived by the sentiment analysis unit 390 and the National Assembly passability (static passability prediction) of each National Assembly bill calculated by the passability calculation unit 320. value) as an input of at least one type of artificial intelligence algorithm, it is possible to derive (obtain) a dynamic passability prediction value of each National Assembly bill through the artificial intelligence algorithm.

Here, at least one type of artificial intelligence algorithm is an artificial intelligence model (deep learning model, machine learning model) that has been pre-learned to output, for example, public opinion information and static passability prediction values as inputs and a corresponding dynamic passability prediction value. model, etc.).

This server 300 uses social data to dynamically predict the passage of each bill in the National Assembly (that is, to derive a predictive value for dynamic passage), and the emotion represented by the pass or non-pass opinion on the bill. By deriving a kind of parameter called analysis data, it is possible to dynamically predict the possibility of passing a bill (a bill) in the plenary session of the National Assembly using the derived parameter (sentiment analysis data). The parameter (sentiment analysis data) at this time is a floating variable, not a fixed variable, and can be continuously updated according to the results of social data collection and analysis. The server 300 uses these parameters to submit the bill to the National Assembly. It is possible to predict the dynamic change value of the legislative impassability that changes from time to time (ie, derive the dynamic passability predicted value).

Meanwhile, in the above description, in explaining the visual representation unit 330, the modeling unit 340, and the error correction unit 350, the visual representation when the server 300 statically predicts the possibility of passing a bill in the plenary session of the National Assembly. The unit 330, the modeling unit 340, and the error correction unit 350 have been illustrated as being applied, but this is only one example to aid understanding of the present application, and is not limited thereto. Even if the contents are omitted below, the description of the visual representation unit 330, the modeling unit 340, and the error correction unit 350 is based on the dynamic prediction of the possibility of passing the bill by the server 300 in the plenary session of the National Assembly. The same can be applied to one description.

As described above, the keyword extraction unit 360 may automatically extract keywords from the contents of each bill. The social data collecting unit 370 may collect social data, such as news, by using the keywords of the extracted bill, store them in the social database 380, and manage them. The sentiment analyzer 390 may perform sentiment analysis to determine whether a text-based opinion of social data stored in the social database 380 is a positive or negative opinion through an algorithm. The dynamic passability prediction unit 391 utilizes the result of the sentiment analysis and the passability computation result from the passability calculation unit 320 (i.e., the static passability predicted value) to dynamically calculate the passageability of each bill in the National Assembly. can do.

That is, the present server 300 can predict the passageability of the bill at the time when the bill is submitted to the National Assembly through the passageability calculation unit 320, and furthermore, after the submission of the bill through the dongtuk passageability prediction unit 391 As time passes, the possibility of passing the bill can also be dynamically predicted by capturing changes in the overall environment surrounding the bill. After the bill is submitted to the National Assembly, the server 300 collects and analyzes opinions or public opinion surrounding the bill based on social data to dynamically predict the passage of the bill.

This system 1 and this server 300 derives a kind of parameter called emotional data represented by opinions on passing or not passing the bill by using data that automatically collects and analyzes a large amount of social data, including news, and , It predicts the possibility of passing the bill in the National Assembly plenary session by conducting continuous learning and testing using artificial intelligence algorithms. It is continuously updated according to the result, and it is possible to predict the dynamic change value of the legislative passability that changes every moment.

This system 1 and this server 300, the amount of social data (buzz) for a specific bill formed on social networks such as news, SNS, blogs, and communities, and positive or negative opinions flowing on social data The possibility of passing a bill in the plenary session of the National Assembly is predicted by considering parameters including trends, etc., and the parameter is continuously and automatically updated as a floating variable rather than a fixed variable to dynamically predict the possibility of passing the bill, depending on whether the bill passes or not. It is possible to contribute to the democratic and open operation of the National Assembly, the representative body of the people, by allowing the people, including affected companies and institutions, to express their opinions and monitor the legislative process more effectively at a lower cost.

In predicting the possibility of a bill passing the National Assembly, static prediction technology alone is insufficient to provide strategic decision-making support services through prediction of the passage possibility of a bill. In other words, the passability of a bill changes every moment according to changes in the environment (public opinion, etc.) surrounding the bill, just like a living organism. There are difficult aspects. In order to solve this problem, the present system 1 and this server 300 quantify the change in public opinion in the passability through static prediction (ie, the static prediction result value derived through the passability calculation unit 320) to obtain the predicted value. By reflecting in, it is possible to provide dynamic predicted results for the passage of the bill.

That is, this system 1 and this server 300 apply sentiment analysis to social data in order to analyze the trend of public opinion surrounding the bill, and consider sentiment analysis data as a parameter, but in particular, a variable variable that is not a fixed variable. By considering it as a result, it is possible to provide dynamic predicted result values that reflect changes in the environment (public opinion) in relation to the possibility of passing the bill in the National Assembly.

As described above, in the present invention, the dynamic passability prediction unit 391 'dynamically predicts passability (ie, dynamic prediction)' means after the existing static prediction is made (ie, the passability calculation unit (320) to read (detect) changes in the passage and possibility of the National Assembly for each National Assembly bill by collecting, tracking, and analyzing changes in the changing environment, public opinion, etc.) that) can be

At this time, after the passability calculation unit 320 calculates the static passability prediction value and the social data collection unit 370 collects and stores the social data, the dynamic passability prediction unit 391 then performs the dynamic prediction. Based on the collected social data, it monitors and analyzes changes in environmental factors, detects outliers through monitoring and analysis, and dynamically predicts passability by dynamically reflecting the detected outliers to the static estimate of passability. can be done

Here, 'detecting an outlier' means to significantly change/affect the existing passability (ie, the passability static prediction value calculated by the passability calculation unit 320) through monitoring and analysis of the collected social data. It can mean generating (significant increase in buzz volume, etc.) or detecting (reading) changes (significant change in positive/negative ratio, etc.) At this time, the setting values (reference value, threshold value) of the outliers can be set or changed in various ways by the user (manager) who manages the present server 300 .

The amount of buzz refers to the number of mentions of a certain topic online, which means to check the reaction of people (the public) to a specific topic (in the present invention, it means a submitted bill) or to determine what It can refer to information used to check where the formation of public opinion on an issue is flowing.

In this way, when an outlier is detected through the analysis of the collected social data, the dynamic passability predictor 391 converts the detected outlier (ie, an outlier representing significant change information in the environment surrounding a specific National Assembly bill information) to a static prediction of passability. By reflecting it in the National Assembly, it is possible to perform a dynamic prediction of the possibility of passing the National Assembly of each National Assembly bill information. Such detected outlier information may mean information on the performance result of the aforementioned sentiment analysis, information on analysis of environmental factors, information on detecting changes in public opinion, and the like.

In addition, this server 300 goes beyond predicting static/dynamic passability using artificial intelligence algorithms, and analyzes information on predicting influence after passage due to the actual passage of the information on the actually passed National Assembly bill, may also provide.

In this case, the influence prediction information may mean prediction information about change information that appears after the National Assembly bill information is actually passed. Such impact prediction information includes, as an example, the pattern of public opinion on the information of the passed National Assembly bill (pattern of public opinion analyzed based on social data), whether or not there is a possibility of abuse, the rate of possibility of abuse (%), examples of abuse cases, and the cost within the bill information of the National Assembly Information such as the difference between the estimate and the budget estimate due to the actual implementation of the relevant National Assembly bill information may be included.

Such influence prediction information is, for example, artificial intelligence algorithms that comprehensively consider information on past National Assembly bills passed in the past, static/dynamic passability prediction information for past National Assembly bill information, and social data collected in relation to past National Assembly bill information. It can be derived by performing learning with The artificial intelligence algorithm may be pre-learned in consideration of various related parameters for deriving the aforementioned influence prediction information.

On the other hand, the server 300 controls the operation of each unit in the server 300 and controls the operation of each of the at least one user terminal 100 interlocked with the server 300 through the network 200 ( 392) may be included.

In this system 1, at least one user terminal 100 may include a member terminal possessed by a member of the National Assembly (member of the National Assembly) capable of proposing (submitting) a bill to the National Assembly and a general user terminal possessed by a general user.

Normally, a legislative bill (bill) can only be submitted with the consent of 10 or more members of the National Assembly. Legislative bills can be submitted in person or online at the National Assembly doctor's office.

Until now, bill proposals have been submitted directly to the Legislation Division after each member's office has printed out the bill as a document and received the signatures of the lawmakers who intend to jointly propose it. After that, when the Bill Division assigns a bill number, the bill is transferred to the online bill information system (National Assembly Security Information System). is disclosed on In the case of the visiting method, visit the National Assembly Legislation Division and submit an application for proposing a legislative bill, a bill proposing a bill, a copy of the signatures of the proponents (10 or more including the representative proponent), and, in the case of a bill involving a budget, a cost estimate along with the bill file. do. In the signatory of the bill, the name of the member of the National Assembly proposing the bill as a representative is written at the top, and the members of the National Assembly co-sponsoring the bill are listed below it. The format (form) for proposing bills in the National Assembly is already established.

A drafting support system has been introduced so that bill proposals can be processed online to solve the inconvenience caused by visiting applications and to improve the efficiency of legislative work (improvement of administrative efficiency). The drafting support system can only be used by insiders of the National Assembly, and it is possible to log in with the ID of a member of the National Assembly. The drafting support system provides functions such as bill submission, joint initiative or online signature for bills, other bill-related information, and instructions for use. After selecting a legislative bill to be submitted to the National Assembly in the drafting support system and completing the process of posting the signatures of co-sponsoring members (at least 10 or more), approval and official documents are sent through the internal electronic document system of the National Assembly, so that they are automatically accepted in the bill section. do.

However, the current drafting support system is rarely used in practice because it has many inconveniences compared to the visit reception, which directly submits bills to the Legislation Division of the National Assembly Secretariat. Illustratively, if any one of the plurality of members (for example, the first member of the legislature) wants to submit a legislative bill (a bill) through the drafting support system in the past, i) input the bill information to be submitted into the drafting support system and ii) Click the registered bill information and then click the Post Name button to receive signatures from other members, iii) At this time, other members log in to the drafting support system to post their names. iv) the first lawmaker enters the drafting support system and inquires about the signature status; 1 After the member selects the registered bill information and clicks the signature closing button for the type of signature publication, the signature publication is closed, vi) When the signature publication deadline is reached, the bill information (legislation information) registered by the first member may be automatically submitted to the Department of Ophthalmology.

That is, in the conventional method of submitting a legislative bill through the drafting support system (online submission method), a member of the legislature who wants to submit a legislative bill must access the drafting support system from time to time to check the status of signatures of other legislators, and the member concerned directly presses the signature close button. The submission of the bill to the Legislation Division was done only by clicking on it, and even when other members signed, it was possible to directly connect to the drafting support system and check the registered bill information before signing. It took a long time, and the process was complicated and inconvenient to a large extent.

In order to solve this problem, the control unit 392 of the present server 300 may provide legislators with a legislative bill reception service so that legislators can more conveniently submit (receive) legislative bills online. The present system 1 may include a plurality of member terminals (not shown) possessed by each of the plurality of members. The description of any one member terminal (for example, the first member terminal) among the plurality of member terminals can be equally applied to the description of each of the plurality of member terminals, even if omitted below. Each of the plurality of clinic terminals may be, for example, a terminal such as a portable mobile phone or a smart phone.

Specifically, in order to provide a legislative bill reception service, the control unit 392 displays a bill (legal bill) information input screen on the screen of the first member terminal when a click is made on the reception button in the first member's terminal possessed by the first member. , the first bill information input by the first member of parliament on the bill information input screen may be received from the first member of the member terminal and stored in an accepted bill DB (not shown) in the server 300 .

The bill information input screen may refer to a screen on which information on a bill (legal bill) that the first lawmaker wants to submit can be input. The bill information input screen may include items for inputting various information required by the bill form, and input items may include, for example, bill type, bill name, reason for proposal, and main contents. According to this, the bill information input by the first member of the bill information input screen may be referred to as the first bill information, and such first bill information may mean information on a bill that the first member of the legislative body intends to submit to the National Assembly. have.

When the first bill information is received from the first clinic terminal, the control unit 392 stores it in the accepted bill DB (not shown) in the server 300 and interlocks with the server 300 through the network 200. A signature request message requesting signature publication for the first bill information is transmitted to the terminals of the plurality of other member terminals (for example, the second member terminal to the nth member terminal) for which message reception is set to allow message reception. can After receiving the signature request message and allowing message reception, members of the member terminals check the contents of the first bill information with their own terminals, and then respond to whether they agree (yes or against) to the motion for the first bill information. It can be performed (transmitting a response message). That is, a response message that can be provided as a response corresponding to the signature request message by the member terminal allowing message reception may be an approval response message or a rejection response message.

The control unit 392 receives a response message corresponding to the signature request message from the member terminals allowing message reception, and when the received plurality of response messages satisfy preset submission conditions, they are submitted at the time when the preset submission conditions are satisfied. After generating the first bill information for bill submission, the generated first bill information for submission is transmitted to the bill department, so that it can be automatically accepted in the bill department.

At this time, the first bill information for submission may be differently referred to as first legislative bill information, first National Assembly bill information, and the like. In addition, the first bill information for submission is reflected (posted) in the first bill information input by the first member of the legislature on the bill information input screen, and the signature of the member of the member terminals allowed to receive the message that sent the affirmative response message as a response message. It may refer to generated bill information. The response message transmitted to the server 300 from the terminal of the member allowing message reception includes information on whether or not to approve (that is, information on whether to agree or disagree), and if yes, the signature (signature) information of the corresponding member in favor. can

The control unit 392 can control the signature request message to be displayed in a pop-up form on the screens of the member terminals that are allowed to receive the message, and through this, the member who receives the signature request message does not require a separate login or user authentication, but the screens of the member terminals. A response to the signature request message can be immediately performed based on the signature request message exposed on the screen.

Here, the preset submission condition means a condition in which the number of affirmative response messages among the plurality of response messages received from the member terminals allowed to receive messages is equal to or greater than the preset number, wherein the preset number is each It may be set differently.

When a lawmaker intends to propose a specific bill (submitting a bill), the motion can be initiated only when there is a certain number of votes in favor. At this time, the number of supporters is set differently for each type of bill. Illustratively, if the bill type is 'consent from the committee or general consent', the consent of two or more members is required, and if the bill type is 'consent of a closed meeting or the proposition of a legislative bill or other bill', the consent of 10 or more members is required. There must be an affirmative vote, and if the type of bill is 'consent to change the agenda, request for attendance by members of the State Council, request for urgent pending issues, request for disciplinary action by a member of the National Assembly', at least 20 members must be in favor. In addition, if the bill type is 'consent to amend the budget bill', there must be at least 50 members in favor.

Therefore, when the bill type of the first bill information is 'legislative proposal', the control unit 392 transmits a signature request message to the terminals of lawmakers allowed to receive messages, and then sends a response message from the terminals of lawmakers allowed to receive messages in response thereto. When receiving, the received response message is analyzed, and the time when 10 affirmative response messages are received as the received response message is recognized as the time when the preset submission condition is met, and the time when 10 affirmative response messages are received First bill information for submission may be generated and transmitted to the bill department.

In addition, the control unit 392 performs the following so that the member terminals allowing message reception can respond more quickly to the signature request message (ie, so that the member terminals allowing message reception can more quickly transmit an approval response message or a rejection response message). can be controlled as If there is no response from the time of transmitting the signature request message to the terminals of the clinics that are allowed to receive the message to the time when a preset time (eg, 1 hour) has elapsed, the control unit 392 determines that the preset time (eg, 1 hour) has elapsed. A response prompt notification message prompting a response to the signature request message may be provided to member terminals allowing message reception at every predetermined prompt notification period (for example, 10 minutes) after the point in time.

When providing a response prompt notification message every time a predefined prompt notification period (eg, 10 minutes) elapses, the control unit 392 may vary the notification type of the response prompt notification message provided, and repeatedly provide the response prompt notification message. It can be provided by changing it to a notification type with higher intensity than The notification types of the response prompt notification message include, for example, a first notification type that is a message type, a second notification type that is a combination of a message and a vibration, a third notification type that is a combination of a message and a ringtone, and a fourth notification type that is a call connection type. can do.

Exemplarily, the control unit 392 confirms that the terminal of the clinic that is allowed to receive the message has not responded until one hour has elapsed (for example, 13:30) after transmitting the signature request message to the terminal of the clinic that is allowed to receive the message. In this case, a response prompting notification message may be provided to the terminal of the doctor who is allowed to receive the message at every predefined prompting notification period (for example, 10 minutes) to prompt the response. At this time, the control unit 392 provides the first notification prompting notification message in the first notification type (message form) to the clinic terminal allowing message reception at 13:30, and if there is no response, the control unit 392 sends a notification message at 13:30. At 40 minutes, the notification message for prompting the second response may be changed to the second notification type (a type in which vibration is generated in the terminals of clinics that are allowed to receive the message in addition to providing the message) and provided. If there is still no response even though the response prompt notification message is provided as the second notification type, the control unit 392 may provide the prompt notification again as a third notification type (message and ring tone combined type). have.

In this way, the control unit 392 transmits the signature request message, and if there is no response within a specific time, a response prompt notification message prompting a response to the signature request message is repeatedly sent every predefined prompt notification period (eg, 10 minutes). It is provided to a reception-permissible clinician terminal, but when a response prompting notification message is repeatedly provided, the notification type may be changed to a notification type with higher recognition strength and provided differently. According to the provision of such a response prompting notification message, the server 300 allows the first member to more quickly submit the bill (legal bill) he/she wants to submit, and provides a signature request message to the member terminal who is allowed to receive the message. It can be provided so that the recognition that has been received is made sure and responds quickly accordingly.

In addition, as described above, when the first bill information is received from the first member terminal, the control unit 392 allows message reception among a plurality of other member terminals (eg, the second member terminal to the nth member member terminal). A signature request message can be transmitted to terminals of members whose messages are allowed to be received.

At this time, it is assumed that the number of member terminals allowed to receive messages is 8, including the second to tenth member terminals. And, in the case of information on the first bill, for example, let's assume that the type of bill is a bill proposal, so that the approval of 10 members is required (ie, the number of supporters must be 10 or more). In this situation, even if all member terminals (i.e., 8 member terminals allowed to receive messages) that have received the signature request message provide a response with an affirmative response message, information on the first bill can be transmitted to the National Assembly using only their responses. It is in a state where it cannot be submitted as a legislative bill. In other words, the information on the first bill can be submitted as a bill only when two more members agree.

In order to solve this problem, the control unit 392 sets a first value (for example, since 10 lawmakers must agree, 10 ) and a second value (for example, 8), which is the number of message receiving member terminals for which message reception is set to allow, and if the second value is equal to or greater than the first value, the second value corresponds to It is possible to transmit a signature request message to member terminals that are allowed to receive messages.

On the other hand, the control unit 392 compares the first value and the second value with each other, and if the second value is smaller than the first value, the difference value between the first value and the second value (for example, 10-8 = 2) is calculated, and thereafter, by a number corresponding to the calculated difference value among the terminals of a plurality of other member terminals (for example, the second member terminal to the nth member terminal) for which message reception is set to not permitted, among member terminals not permitted to receive messages. Selecting a member terminal at random (that is, selecting a member terminal that is not allowed to receive two messages, and using the ARS function for each randomly selected member terminal to provide signature request information requesting signature publication for the first bill information) If the member terminal receiving the signature request guide information (ARS) responds with acceptance, the control unit 392 transmits a signature request message to the member terminal that has responded to the acceptance and sends a response to the corresponding member terminal can be received from

In addition, the control unit 392 transmits response messages (ie, all of the member terminals that receive the signature request message from the server 300) to all of the member terminals (which may include the member terminals allowed to receive messages and the member terminals randomly selected). response message), if the third value, which is the number of affirmative response messages in the entire response message, is smaller than the first value (i.e., 10), which is the minimum required number of supporters required in relation to the first bill information, Among the plurality of other member terminals (for example, the second member terminal to the n-th member terminal), the third value and the first value are applied to the remaining member terminals excluding the member terminals that have received the signature request message from the server 300. After randomly additionally selecting as many member terminals as the number corresponding to the difference value between, signature request guide information for requesting signature publication for the first bill information may be provided to each of the additionally selected member terminals through the ARS function. The controller 392 may repeatedly perform this process (additional selection process) until the third value becomes the same value as the first value. That is, the control unit 392 may repeatedly perform a process of additionally selecting at random until the number of affirmative response messages is equal to the minimum necessary number of affirmative voters.

This server 300 transmits a signature request message to the member terminals that allow message reception, but if an additional signature request is required, additional signature requests are sent only to a minimum number of randomly extracted member terminals so that signatures can be made. As such, it is possible to promptly submit a bill based on the signatures in favor of the minimum required number of supporters.

According to this, when the first lawmaker wants to submit a legislative bill (first bill information) online, this server 300 simply inputs the first bill information into this server 300, and the first bill information Information on the first bill for submission can be automatically transmitted (submitted) to the bills section as bill information with the signature information of the member in favor of the bill being automatically reflected.

That is, in this server 300, in order to submit a legislative bill, the first member of the legislative body separately visits several members and obtains signatures in favor, or accesses the drafting support system from time to time to check the signature status of other legislators and determines When the signature is no longer required, it is automatically submitted when the preset submission conditions are met by simply entering the first bill information into the server 300 without the need to take actions such as clicking the signature close button. Since the information on the first bill can be transmitted (submitted) to the Legislative Division, it is possible to quickly and easily submit bills (legal proposals), thereby improving the convenience of users (members).

In addition, as mentioned above, when the first lawmaker wishes to submit a bill to the National Assembly, automatic submission of the bill becomes possible by inputting the first bill information on the bill information input screen provided by the server 300. can

At this time, the control unit 392 may provide an input completion button in the first area of the bill information input screen, and when the first member clicks on the input completion button, the first bill information input by the first member A signature request message may be transmitted at the same time as storing in a reception bill DB (not shown).

Also, the controller 392 may provide a list of similar bills to the second area of the bill information input screen. The similar bill list may refer to a list of similar bill information having similarity with the first bill information input by the first member of parliament on the bill information input screen equal to or higher than a preset similarity value (for example, 90%).

At this time, the similar bill information included in the list may be, for example, information on at least some of the information on a plurality of National Assembly bills stored in the National Assembly bill information database 500, but is not limited thereto, and as another example, overseas bills abroad. It may be at least part of the information of a plurality of bills stored in the overseas bill DB in association with the DB. The overseas legislative bill DB may refer to a DB in which information on legislative bills of various foreign countries, such as the United States and Japan, is stored. Hereinafter, for convenience of explanation, information on similar bills will be described as being at least a part of information on a plurality of bills of the National Assembly stored in the National Assembly bill information database 500 as an example.

The control unit 392 may perform similarity comparison between the first bill information and the plurality of pieces of National Assembly legislative information stored in the National Assembly legislative bill information database 500 in order to provide a list of similar bills in the second area of the bill information input screen. . At this time, when comparing the similarity between the two bill information (ie, the first bill information and the information on a plurality of National Assembly bill information), various similarity comparison techniques (eg, similarity between documents) that have been previously known or developed in the future analytical techniques) may be applied. The degree of similarity may be analyzed based on, for example, the number of matching words.

After the first member inputs the first bill information before clicking the input completion button, the control unit 392 converts information on the National Assembly bill information (similar bill) having a high degree of similarity to the input bill information based on the input first bill information. After being identified in the information database 500, a similar bill list, which is a list of identified similar bills, may be displayed in the second area. The first legislator has the list of similar bills displayed in the second area, among the information on multiple domestic/overseas national assembly bills that have been submitted in the past, to check what kinds of information are examples of bills similar to the bill he or she is currently trying to submit. It could be possible.

Also, as an example, assume that information on five similar bills is included as similar bill information in the list of similar bills. When the list of similar bills is displayed in the second area, bill name information may be displayed as information of each similar bill, for example.

At this time, the control unit 392, when a click input (click input within 1 second) of the first member of the parliament is made on the first similar bill information in the similar bill list displayed in the second area, the information displayed on the first member terminal Operation of the terminal of the first member of parliament may be controlled so that the bill information input screen is changed to a detailed information display screen of the first similar bill information. That is, when the first lawmaker clicks the first similar bill information, the display screen of detailed information of the first similar bill information may be displayed.

If the control unit 392 detects that the first member pressed input for a preset pressing time (eg, 3 seconds) or longer with respect to the first similar bill information in the similar bill list displayed in the second area (eg, 3 seconds), If it is pressed for 3 seconds and then released), it is recognized as a signal requesting that the display screen of the detailed information of the first similar bill information be displayed in a pop-up form (ie, the screen of the member terminal displays the detailed information of the first similar bill information) Instead of going to the page provided, it is recognized as a signal requesting that the display screen of the detailed information of the first similar bill information be displayed in an overlap form in a certain area on the screen of the current member terminal), The display screen of the detailed information may be displayed in a pop-up form on the current screen of the first clinic terminal (overlapping one area on the bill information input screen).

In addition, a turn off button may be provided in one area of the overlap display area so that the screen of detailed information displayed overlapping is not exposed on the screen (ie, not displayed on the screen of the first clinic terminal). When the first member clicks the off button, the detailed information screen overlapping and displayed on the first member terminal may disappear from the screen.

In this way, the control unit 392, with respect to each of the plurality of similar legislative bill information in the similar bill list, when the first member presses and releases the information for a preset pressing time (for example, 3 seconds) or longer, displays the corresponding similar legislative bill pressed for a preset pressing time or longer. Detailed information of the information may be displayed in an overlapping form in a predetermined area on the screen of the current first clinic terminal. This server 300 can compare similar bill information similar to the first member information (law bill) that the first member wants to submit before the first member inputs the first member information and completes registration/storage. In particular, when comparing, it is possible to efficiently compare bills without having to go back and forth (ie, without the need for hassle), such as going to a detailed page of information on a specific similar bill and then returning to the previous page.

In addition, the control unit 392 may expose a list of a plurality of pieces of National Assembly bill information stored in the National Assembly bill information database 500 on a screen of a user terminal (including a member terminal or a general user terminal). The control unit 392 may provide a search function enabling a user to search information on the National Assembly bill desired by the user in the list. At this time, in order to provide a search function, the control unit 392 overlaps and displays a keyword search icon on one area of the screen of the user terminal when displaying the list, and matches a keyword input through the keyword input unit within the keyword search icon. A search result obtained by searching for a keyword in one search mode selected by a user among a plurality of search modes may be displayed on the screen. The keyword input unit may consist of one column.

Here, the plurality of search modes may include a first search mode, a second search mode, a third search mode, and a fourth search mode. The first search mode may be a mode for extracting information on National Assembly bills including spaces and including information input to the keyword input unit and matching information (same information) and displaying them in a list. For example, if information input to the keyword input unit is 'installation of sports facilities', only a list of information on the National Assembly bill including 'installation of sports facilities' may be provided as a search result.

In the second search mode, when spaces are included in the information input to the keyword input unit, words before and after are recognized as individual words based on the spaces, but information on the National Assembly bill including at least one of the individually recognized words is extracted and listed. It may be a mode indicated by . For example, if the information entered in the keyword input unit is 'installation of sports facilities', information on the National Assembly bill including 'sports facility', information on the National Assembly bill including 'installation', and information on the National Assembly bill including 'installation of sports facilities' will appear in the search results. Can be provided as a list.

In the third search mode, when spaces are included in the information entered into the keyword input unit, words in front and behind are recognized as individual words based on the spaces, and individual recognized words It may be a mode that extracts information on the bills of the National Assembly described in one sentence and displays them in a list. Here, a single sentence may be a sentence classified based on a 'period (.)', or if a paragraph is converted to an enter without a period, the corresponding paragraph without a period may be regarded as a sentence. For example, if the information entered in the keyword input unit is 'installation of sports facilities', information on the National Assembly bill including the word 'sports facility' and information on the National Assembly bill including the sentence 'where to install sports facilities' It can be provided as a list of search results.

In the fourth search mode, when spaces are included in the information entered into the keyword input unit, words in front and behind are recognized as individual words based on spacing, but the text information entered in each National Assembly bill information is linked to each National Assembly bill information It may be a mode in which individual recognized words are extracted and displayed in a list, targeting registered attachments, in which information on the National Assembly bill is described in one sentence. If the information entered in the keyword input unit is 'installation of sports facilities', the words 'sports facilities' or 'installation' are not included in the text information entered in the bill information of the 1st National Assembly, but in the information of the bill of the 1st National Assembly If the sentence 'Where to install sports facilities~' is included in the attached pdf attachment, information on the 1st National Assembly bill may be provided as a list of search results.

In general, when performing a keyword (word) search on the web or on a document such as a word file, for example, if you enter information such as 'installation of sports facilities' with spaces, only the search for the word "installation of sports facilities" is possible. did

As another example, conventionally, when searching for specific information, as shown in FIG. 4A, if each information is input for each of a plurality of items, only information including all the inputted information (ie, all inputted information) Only information matching information) was provided as a search result. In addition, conventionally, when a post is uploaded with a file such as a pdf file or a Hangul file attached, when a keyword search is performed for the uploaded information, the search is generally performed only for the text in the post, It was an impossible structure to search within the attached file uploaded with .

As such, the conventional search technology lacks diversity in search methods, and thus has limitations in effectively searching for information desired by the user. In order to solve this problem, the server 300 determines whether a space is included in the information in relation to the single keyword information input into the keyword input unit, and the user selects a plurality of search modes (first search mode to fourth search mode). ), different search results can be provided depending on which search mode is selected, etc., so that the user (clinician, etc.) can effectively search for desired information.

5 is a diagram showing a process of transmitting and receiving data between components included in the system 1 for providing a dynamic prediction service for the possibility of passing a bill in the plenary session of the National Assembly using social data analysis of FIG. 1 according to an embodiment of the present invention. . Hereinafter, an example of a process of transmitting and receiving data between each component will be described through FIG. 5, but the present application is not limited to such an embodiment, and according to various embodiments described above, It is obvious to those skilled in the art that a process of transmitting and receiving data may be changed.

Referring to FIG. 5, the system 1 for providing a dynamic prediction service for the possibility of passing a bill in the plenary session of the National Assembly using social data analysis automatically collects information on the National Assembly bill in real time from the National Assembly bill information system in the open API bill information collection server (S5100 ), and stores the information on the National Assembly bill that is automatically collected in real time from the open API bill information collection server in the National Assembly bill information database (S5200).

Then, after the coding unit of the prediction service providing server automatically codes the National Assembly bill information stored in the National Assembly bill information database in real time (S5300), the calculation unit of the prediction service providing server, based on the automatically coded result in the coding unit The possibility of passing the National Assembly (static prediction value) of each bill of the National Assembly is calculated (S5400). In addition, after step S5400, the dynamic passability prediction unit analyzes the environmental factors affecting the passability of the bill submitted to the National Assembly in relation to the National Assembly bill information stored in the National Assembly bill information database, so that each National Assembly bill is capable of passing the National Assembly. can dynamically predict (calculate a dynamic predicted value).

The order between the above-described steps (S5100 to S5400) is only an example, and is not limited thereto. That is, the order of the above-described steps (S5100 to S5400) may be mutually changed, and some of the steps may be simultaneously executed or deleted.

Matters that have not been described in the method for providing a service for dynamically predicting the possibility of passing a bill in the plenary session of the National Assembly using the analysis of social data in FIG. Since the content described for the provision system 1 and the server 300 is the same as or easily inferred from the description, the following description will be omitted.

The method for providing a service for dynamically predicting the passability of a bill in the plenary session of the National Assembly using social data analysis according to an embodiment described with reference to FIG. It can also be implemented in the form of a medium. Computer readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, computer readable media may include all computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

The above-described method for providing a dynamic prediction service for the possibility of passing a legislative bill in the plenary session of the National Assembly using social data analysis according to an embodiment of the present invention is an application basically installed in a terminal (this is a program included in a platform or operating system that is basically installed in the terminal). It may be executed by an application store server, an application, or an application (ie, a program) installed directly in the master terminal by a user through an application providing server such as a web server related to the service. In this sense, the above-described method for providing a service for dynamically predicting the possibility of passing a bill at the plenary session of the National Assembly using social data analysis according to an embodiment of the present invention is implemented as an application (ie, a program) that is basically installed in a terminal or directly installed by a user. and can be recorded on a computer-readable recording medium such as a terminal.

The above description of the present invention is for illustrative purposes, and those skilled in the art can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts should be construed as being included in the scope of the present invention. do.

1: System for providing dynamic prediction service for legislative bill passing possibility in plenary session of National Assembly using social data analysis
300: prediction service providing server
310: coding unit 320: passability calculation unit
330: visual expression unit 340: modeling unit
350: error correction unit 360: keyword extraction unit
370: social data collection unit 380: social database
390: Sentiment Analysis Unit 391: Dynamic Passability Prediction Unit
392: control unit

Claims (3)

As a system for providing a dynamic prediction service for the passage of a bill in the plenary session of the National Assembly using social data analysis,
An open API bill information collection server that automatically collects information on the National Assembly bill in real time from the National Assembly bill information system;
a National Assembly bill information database for storing the National Assembly bill information automatically collected in real time from the open API bill information collection server; and
A coding unit for automatically coding the National Assembly legislative bill information stored in the National Assembly bill information database in real time, a passability calculation unit for calculating the passage possibility of each National Assembly bill based on the result automatically coded in the coding unit, and the stored National Assembly bill A prediction service providing server including a dynamic passability prediction unit that dynamically predicts the passage possibility of each bill of the National Assembly by analyzing environmental factors affecting the passage of the bill submitted to the National Assembly in relation to information,
The environmental factors include public opinion information identified based on social data collected in relation to the submitted bill,
The prediction service providing server further includes a control unit for controlling the operation of each of at least one user terminal interlocked through a network,
The at least one user terminal includes a member terminal possessed by a member of the National Assembly capable of proposing a bill to the National Assembly and a general user terminal possessed by a general user,
The control unit,
A list of the plurality of National Assembly bill information stored in the National Assembly bill information database is exposed on the screen of the user terminal, and a search function is provided that enables the user to search for the National Assembly bill information desired by the user in the list, but to provide the search function When displaying the list, a keyword search icon is overlapped and displayed on one area of the screen of the user terminal, and a matching keyword that matches the keyword entered through the keyword input unit in the keyword search icon is searched in any one selected by the user from among a plurality of search modes. The search results searched by mode are displayed on the screen,
The plurality of search modes include a first search mode, a second search mode, a third search mode, and a fourth search mode,
The first search mode is a mode for extracting and displaying information in a list of National Assembly legislative bills including information that is entirely consistent with information entered in the keyword input unit, including spaces,
In the second search mode, when spaces are included in the information input to the keyword input unit, words in front and behind are recognized as individual words based on spaces, and information on the National Assembly bill including at least one of the individually recognized words is extracted. This is a list display mode.
In the third search mode, when spaces are included in the information input to the keyword input unit, words in front and behind are recognized as individual words based on the spaces, but target text information entered in each National Assembly bill information, individually recognized This is a mode that extracts the information of the National Assembly bills in which words are written in one sentence and displays them in a list.
In the fourth search mode, when spaces are included in the information input to the keyword input unit, words in front and behind are recognized as individual words based on spaces, and text information entered in each National Assembly bill information is linked to each National Assembly bill information. Dynamic prediction of the possibility of passing the National Assembly plenary session of a bill using social data analysis, which is a mode that extracts information on the National Assembly bill in which individually recognized words are written in one sentence and displays it as a list, targeting the attached file that has been registered and registered. service delivery system. According to claim 1,
The prediction service providing server,
a keyword extraction unit for extracting a keyword for a bill from the contents of the stored bill information of the National Assembly; a social data collection unit that collects social data related to the extracted keywords of the legislative bill from a social network service providing server and stores them in a social database; and a sentiment analysis unit deriving sentiment analysis data corresponding to the social data as the public opinion information by performing sentiment analysis on the social data stored in the social database.
A system for providing a dynamic prediction service for the possibility of passing a legislative bill in the plenary session of the National Assembly using social data analysis further comprising. According to claim 1,
The prediction service providing server,
Pre-set information among the National Assembly bill information stored in the National Assembly bill information database is selected as at least one factor affecting the passageability of the National Assembly plenary session, and the relationship between the at least one factor and the passageability is determined by at least one artificial intelligence A modeling unit that builds a prediction model for passing the plenary session of the National Assembly by learning and testing with algorithms and machine learning;
A system for providing a dynamic prediction service for the possibility of passing a legislative bill in the plenary session of the National Assembly using social data analysis further comprising.

Images