WO2019103205A1 - System for predicting location-specific crime occurrence probability on basis of machine learning algorithm by using crime-related big data - Google Patents

System for predicting location-specific crime occurrence probability on basis of machine learning algorithm by using crime-related big data Download PDF

Info

Publication number
WO2019103205A1
WO2019103205A1 PCT/KR2017/013563 KR2017013563W WO2019103205A1 WO 2019103205 A1 WO2019103205 A1 WO 2019103205A1 KR 2017013563 W KR2017013563 W KR 2017013563W WO 2019103205 A1 WO2019103205 A1 WO 2019103205A1
Authority
WO
WIPO (PCT)
Prior art keywords
crime
data
pattern
item
risk index
Prior art date
Application number
PCT/KR2017/013563
Other languages
French (fr)
Korean (ko)
Inventor
권민구
Original Assignee
주식회사 웬즈데이에잇피엠
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 주식회사 웬즈데이에잇피엠 filed Critical 주식회사 웬즈데이에잇피엠
Priority to PCT/KR2017/013563 priority Critical patent/WO2019103205A1/en
Publication of WO2019103205A1 publication Critical patent/WO2019103205A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/04—Inference or reasoning models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N99/00—Subject matter not provided for in other groups of this subclass
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services

Definitions

  • the present invention relates to predicting a crime occurrence probability. More specifically, the present invention relates to predicting a crime occurrence probability based on a machine learning algorithm.
  • Big data is data generated in a digital environment, which is large in scale, has a short generation cycle, and is large-scale data including various types of data such as text and image data as well as numerical data. Big Data has emerged with increasing data due to the development of various sensors and the Internet. As computers and processing technologies develop, analysis based on big data generated in the digital environment has increased the possibility of discovering new perspectives and laws about changes in diseases and social phenomena. Therefore, Big Data can be used to solve social problems, especially in crime prevention and investigation.
  • Big Data technology can analyze historical crime data and identify patterns that can predict places and times of high crime risk.
  • crime location and type can be provided in detail around the address, to identify and prevent specific crimes, and to prevent crime by prioritizing police personnel in areas of high crime incidence.
  • the technical object of the present invention is to analyze a pattern of a condition in which a crime occurs using crime-related big data, and to provide information about a crime pattern of a crime to a user.
  • the present invention has an object to precisely measure the position of a user using a fingerprint technique and to provide dangerous crime pattern information on the measured user's position to the user.
  • crime occurrence patterns can be identified and predicted by quantifying the crime occurrence probability according to the crime occurrence classification standard as a risk index.
  • accurate location of the user can be grasped and crime occurrence pattern information optimized for the user can be provided.
  • FIG. 1 is a block diagram illustrating a system for predicting a crime according to an embodiment of the present invention.
  • the crime risk index is calculated by multiplying the maximum value and the minimum value of the number of occurrences of the crime by each item in the category of the crime occurrence, Calculated based on;
  • the crime risk index is calculated on the basis of the ratio of the number of crime incidents of the item to the difference between the maximum value and the minimum value,
  • the crime pattern is a combination of items selected in descending order of the crime risk index in each of the classification criteria
  • the step of collecting the crime data may include receiving formatted data expressed in the form of the number of occurrences of crime for each item of the classification criteria; And acquiring crime related unstructured data using at least one of news data, social network service (SNS) data, and web data,
  • SNS social network service
  • a crime prediction system includes a server 100 and a user terminal 200.
  • the server 100 of the present invention collects crime data and generates and analyzes a crime pattern to provide the data to a user.
  • the server 100 includes a communication unit 110, a data collecting unit 120, a data processing unit 130, a crime pattern generating unit 140 and a big data DB 150.
  • the communication unit 110 can transmit / receive data to / from an external terminal or a network using various wired or wireless communication systems.
  • the available wireless communication methods include Wi-Fi, Long Term Evolution (LTE), Bluetooth, Near Field Communication (NFC), Zigbee, infrared communication, and the like.
  • the data collecting unit 120 collects the crime-related data through the communication unit 110.
  • the crime data may indicate the number of crime occurrences counted according to predetermined classification criteria.
  • the data collection unit 120 may collect crime-related regular data and unstructured data together.
  • Formal data refers to crime data expressed in the form of the number of occurrences of crime for each item in the classification standard, and can be obtained from public authorities such as the Supreme Prosecutor's Office.
  • the atypical data is acquired using at least one of news data, social network service (SNS) data, and web data.
  • the data collecting unit 120 may extract detailed information on a specific crime event from the news data or the like, and collects crime data using the extracted information.
  • the data collection unit 120 may further acquire location information of the user, current time information, and the like.
  • the collected location information of the user, current time information, and the like are transmitted to the data processing unit 130.
  • the data processing unit 130 calculates a crime risk index for each category of classification criteria based on the collected crime data.
  • the crime risk index can be calculated on the basis of the maximum value and the minimum value of the number of crimes of the item, and the number of crimes of each item in the classification standard to which the item belongs.
  • the data processing unit 130 may process the data in the server 100 and provide the processed data or the data stored in the big data DB 150 to the user through the communication unit 110.
  • the crime pattern generator 140 generates a crime pattern based on the calculated crime risk index. More specifically, the crime pattern of the present invention represents information on a combination of items selected for each classification criterion, and can be generated based on the crime risk index. A detailed description thereof will be given later.
  • the big data DB 140 stores various digital data, and may store information such as crime data and crime patterns according to an embodiment of the present invention. According to one embodiment, the big data DB 140 may designate a pattern in which the sum of the crime risk indices for each item out of the crime patterns generated by the crime pattern generator 140 exceeds a reference value as a dangerous crime pattern .
  • the big data DB 140 can be implemented through various digital data storage media such as a hard disk drive, a flash memory, a random access memory (RAM), and a solid state drive (SSD).
  • the user terminal 200 may receive the dangerous crime pattern information from the server 100 and output it as video and / or audio information.
  • the user terminal 200 may include various types of digital devices such as a smart phone, a desktop, a laptop, a tablet PC, a personal digital assistant (PDA), a smart watch, and a head mounted display (HMD).
  • the user terminal 200 may include a communication unit 210, an output unit 220, and a control unit 230.
  • the communication unit 210 can transmit / receive data to / from the server or the network using various wired or wireless communication systems.
  • the available wireless communication methods include Wi-Fi, Long Term Evolution (LTE), Bluetooth, Near Field Communication (NFC), Zigbee, and infrared communication.
  • the output unit 220 may include various types of video and / or audio information output means, such as a display unit, a speaker, and the like.
  • the output unit 220 can output various kinds of information in the form of video and / or audio based on the control of the control unit 230.
  • the control unit 230 may control the operation of each unit of the user terminal 200 and may process data in the user terminal 200. In addition, the control unit 230 can control data transmission / reception between the respective units of the user terminal 200. [ The control unit 230 of the user terminal 200 may control the output unit 220 to output the dangerous crime pattern information received from the server 100. [ Alternatively, the control unit 230 of the user terminal 200 may output an alarm signal based on the dangerous crime pattern information through the output unit 220.
  • the user terminal 200 can measure the position of the corresponding terminal using the fingerprint technique. More specifically, the user terminal 200 can measure the intensity of a communication signal received from a plurality of access points (APs), and position information of the corresponding terminal can be obtained based on the intensity of the measured signal. The user terminal 200 may transmit the obtained location information of the corresponding terminal to the server 100.
  • APs access points
  • the number of crime incidents can be classified according to various classification criteria, for example, according to type of crime, area of crime, location of crime, time of crime occurrence, and the like.
  • Each classification criterion can be composed of at least one item.
  • the type of crime (C) can be composed of different kinds of crime items (C1, C2, C3, ...), which can be organized according to crime classification criteria of the Supreme Prosecutors' And may be uniquely set in the system of the invention.
  • the crime data shows information on the number of crimes according to each crime type item (C1, C2, C3, ).
  • the crime occurrence area (R) can be composed of each item (R1, R2, R3, ...) classified according to the geographical division of the crime occurrence point, for example, More precisely, it may be divided into each phrase.
  • the place of crime (S) can be composed of each item (S1, S2, S3, ...) classified according to the type of the crime spot. For example, apartment, single house, highway, , Units, offices, schools, warehouses, subways, public transportation, and the like.
  • Each item (T1, T2, T3, ...) of the crime occurrence time (T) may be configured as a predetermined time unit, or may be configured as an arbitrary time zone such as dawn, morning, day, morning, afternoon, evening, .
  • the crime classification standard may include a day of crime occurrence (D).
  • Each item (D1, D2, D3, ...) of the day of crime occurrence (D) Monday, ..., Saturday.
  • the crime pattern generator 140 calculates the crime index INDn of the crime occurrence area R according to the crime type C, the item Rn having the high INDRn value, (Dn) of high crime index (INDDn) and high crime index (INDSn) (Sn) among high crime index (Tn), crime day It is possible to generate the crime pattern P (R, T, D, S).
  • the server 100 of the present invention determines whether or not the generated crime risk index sum value INDRTDS exceeds the reference value and if the crime index sum value INCDTDS exceeds the reference value, the server 100 stores the crime pattern as a risky crime pattern in the big data DB 150 do. Accordingly, the big data DB 150 can store information on crime patterns in which the sum of the crime risk indices is larger than the reference value.
  • the crime pattern generator 140 may generate another crime pattern by changing items for each classification criterion, and calculate the sum of crime risk indexes for each item of the crime pattern.
  • the crime pattern generation unit 140 checks whether the corresponding pattern is a crime pattern generated at the time of generating the crime pattern, thereby preventing duplicate crime patterns from being generated. In order to generate different crime patterns, items can be selected in descending order of crime risk index in each classification criterion. If the sum of the crime risk indices for each item of the crime pattern exceeds the reference value, the crime pattern is stored in the big data DB 150 as a dangerous crime pattern.
  • the server 100 of the present invention can provide crime prediction information to a user based on crime pattern information stored in the DB 150, in particular, dangerous crime pattern information and the like.
  • the user terminal 200 can estimate the position of a corresponding terminal by measuring the strength of a signal received from a plurality of APs (Access Points) 50.
  • the user terminal 200 may receive a probe response signal or various signals from the APs 50a, 50b, and 50c, and the received AP signal may include location information of the corresponding AP.
  • the user terminal 200 can acquire position information of the corresponding device using the triangulation method based on the position information of each of the plurality of received AP signals, the intensity information of the received signal, and the like.
  • a space for estimating the location of the user is set in the form of nxm grid and a plurality of APs AP1 (1, 2), ..., P (n, m) , AP2, ..., APk) are measured.
  • the AP signal is measured multiple times at each point, and the average value of the measured values can be stored in the DB.
  • the DB is constructed as described above, it is possible to provide location information of a corresponding terminal corresponding to a combination of strengths of a plurality of AP signals received at the user terminal 200.
  • the user terminal 200 measures the intensity of a signal received from a plurality of APs, compares the intensity of the measured signal with a pre-measured value at each point stored in the DB , It is possible to determine the position where the error of the received signal strength is the smallest as the position of the corresponding terminal.
  • the user terminal 200 may acquire the location information of the user using the GPS signal, the Bluetooth beacon signal, or a combination thereof as well as the AP signal.
  • the location information of the user terminal 200 is obtained, the location information of the terminal is transmitted to the server 100.
  • the server 100 receiving the location information can search for a dangerous crime pattern corresponding to the location information of the corresponding user terminal 200 among the dangerous crime patterns stored in the big data DB 150. That is, the server 100 searches for a dangerous crime pattern having a crime occurrence area / crime occurrence place item matched with the position information of the corresponding user terminal 200.
  • the server 100 of the present invention can obtain current time information.
  • the server 100 may receive the current time information by mounting the GPS module and may receive the current time information together with the location information of the terminal from the user terminal 200 according to another embodiment.
  • the server 100 receiving the current time information can search for a dangerous crime pattern having the crime occurrence time / crime occurrence day item corresponding to the current time among the dangerous crime patterns stored in the big data DB 150.
  • the server 100 acquires the location information of the user terminal 200 together with the current time information, the location information of the crime occurrence / crime occurrence place / crime occurrence time / crime occurrence day Risk crime patterns with items can be retrieved.
  • the server 100 may transmit the detected dangerous crime pattern information to the corresponding user terminal 200.
  • the user terminal 200 receiving the dangerous crime pattern information from the server 100 outputs the information through the output unit 220 with the image and / or voice.
  • the server collects the crime data counted according to predetermined classification criteria (S110).
  • the above-mentioned crime data represents the number of crimes occurring according to each item in the classification standard.
  • the classification criteria may include a crime type, a crime occurrence area, a crime occurrence place, a crime occurrence time, a crime occurrence day, and the like.
  • the server calculates the crime risk index for each item based on the crime data (S120).
  • the crime risk index may be proportional to the number of crimes in the item.
  • the crime risk index can be calculated based on the maximum value and the minimum value of the number of occurrences of crime for each item in the classification standard to which the item belongs as well as the number of crime occurrence of the item. That is, the crime risk index is calculated based on the ratio of the number of crimes of the item to the difference between the maximum value and the minimum value.
  • the server generates a crime pattern for a combination of items selected for each classification criterion based on the calculated crime risk index (S130).
  • the crime pattern is a combination of items selected in descending order of crime risk index in each classification standard.
  • the server determines whether the sum of the crime risk indices of each item of the generated crime pattern exceeds the reference value (S140). If the sum of the crime risk index exceeds the reference value, the server stores the crime pattern in the big data DB as a crime pattern (S150). The server returns to step S130 and changes the item of each classification criterion to generate another crime pattern and repeats steps S140 and S150 for the crime pattern. At this time, the server confirms whether or not the pattern is the same as the previously generated crime pattern, thereby preventing duplicate crime patterns from being generated.
  • crime occurrence patterns can be identified and predicted by quantifying the crime occurrence probability according to the crime occurrence classification standard as a risk index.
  • accurate location of the user can be grasped and crime occurrence pattern information optimized for the user can be provided.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Tourism & Hospitality (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • General Business, Economics & Management (AREA)
  • Primary Health Care (AREA)
  • Marketing (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Educational Administration (AREA)
  • Development Economics (AREA)
  • Artificial Intelligence (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Alarm Systems (AREA)

Abstract

The present invention relates to a crime prediction method and a crime prediction system which use big data-based crime pattern analysis and, more specifically, to a crime prediction method and system for analyzing a pattern for a combination of conditions, in which crime occurrence possibility is high, and providing the same to a user. To this end, the present invention comprises the steps of: collecting pieces of crime data, counted according to preset classification criterion; calculating a crime risk index for each item on the basis of the crime data; generating, on the basis of the calculated crime risk index, a crime pattern for a combination of items selected according to classification criterion; determining whether an aggregate value of crime risk indexes for each item of the generated crime pattern exceeds a reference value; and storing the corresponding crime pattern as a dangerous crime pattern when the aggregate value exceeds the reference value.

Description

범죄 관련 빅데이터를 활용한 미선러닝 알고리즘 기반 위치별 범죄 발생 확률 예측 시스템Crime-Occurrence Prediction System by Location Based on Crime-related Big Data
본 발명은 범죄 발생 확률을 예측함에 관한 것이다. 보다 자세하게는 머신러닝 알고리즘을 기반으로 범죄 발생 확률을 예측함에 관한 것이다.The present invention relates to predicting a crime occurrence probability. More specifically, the present invention relates to predicting a crime occurrence probability based on a machine learning algorithm.
빅데이터란 디지털 환경에서 생성되는 데이터로 그 규모가 방대하고, 생성 주기가 짧으며, 수치 데이터뿐 아니 라 문자와 영상 데이터 등 다양한 형태를 포함하는 대규모의 데이터를 말한다. 빅데이터는 각종 센서와 인터넷의 발달로 데이터가 늘어나면서 등장하게 되었다. 컴퓨터 및 처리기술이 발달함에 따라 디지털 환경에서 생성 되는 빅데이터를 기반으로 분석을 수행하게 되면 질병이나 사회현상의 변화 등에 관한 새로운 시각이나 법칙을 발견할 가능성이 커지게 되었다. 따라서, 빅데이터는 사회적인 문제를 해결하는데 사용될 수 있는데, 특히 범 죄 예방과 수사에 적극적으로 활용될 수 있다.Big data is data generated in a digital environment, which is large in scale, has a short generation cycle, and is large-scale data including various types of data such as text and image data as well as numerical data. Big Data has emerged with increasing data due to the development of various sensors and the Internet. As computers and processing technologies develop, analysis based on big data generated in the digital environment has increased the possibility of discovering new perspectives and laws about changes in diseases and social phenomena. Therefore, Big Data can be used to solve social problems, especially in crime prevention and investigation.
예를 들어, 빅데이터 기술을 이용하면 과거 범죄 데이터들을 분석하고 패턴을 파악함으로써 범죄발생 위험이 높은 장소와 시간을 예측할 수 있다. 또한, 범죄위치와 유형 등을 주소지 주변으로 상세하게 제공하여, 구체적인 범죄에 대한 정보를 파악하고 예방하거나 범죄 발생률이 높은 지역에 경찰 인력을 우선 배치하여 범죄를 예방할 수 있다.For example, using Big Data technology can analyze historical crime data and identify patterns that can predict places and times of high crime risk. In addition, crime location and type can be provided in detail around the address, to identify and prevent specific crimes, and to prevent crime by prioritizing police personnel in areas of high crime incidence.
[선행기술문헌][Prior Art Literature]
[특허문헌][Patent Literature]
국내공개특허 10-2016-0104223호Korean Patent Publication No. 10-2016-0104223
본 발명의 기술적 과제는 범죄 관련 빅데이터를 이용하여 범죄가 발생하는 조건에 대한 패턴을 분석하고, 범죄 발생 가능성이 높은 위험 범죄 패턴에 대한 정보를 사용자에게 제공하기 위한 목적을 가지고 있다. 또한, 본 발명은 핑거 프린트 기법을 이용하여 사용자의 위치를 정밀하게 측정하고, 측정된 사용자의 위치에 대한 위험 범죄 패턴 정보를 해당 사용자에게 제공하기 위한 목적을 가지고 있다.The technical object of the present invention is to analyze a pattern of a condition in which a crime occurs using crime-related big data, and to provide information about a crime pattern of a crime to a user. In addition, the present invention has an object to precisely measure the position of a user using a fingerprint technique and to provide dangerous crime pattern information on the measured user's position to the user.
본 발명의 일 양태에 따르면, 기 설정된 분류 기준 별로 계수된 범죄 데이터를 수집하는 단계, 상기 범죄 데이터는 해당 분류 기준에서의 각 항목에 따른 범죄 발생 횟수를 나타냄; 상기 범죄 데이터에 기초하여 상기 각 항목별 범죄 위험 지수를 산출하는 단계, 상기 범죄 위험 지수는 해당 항목의 범죄 발생 횟수, 해당 항목이 속한 분류 기준에 서의 각 항목 별 범죄 발생 횟수의 최대값 및 최소값에 기초하여 산출됨; 상기 산출된 범죄 위험 지수에 기초하여, 상기 각 분류 기준 별로 선택된 항목의 조합에 대한 범죄 패턴을 생성하는 단계; 상기 생성된 범죄 패턴의 각 항목별 범죄 위험 지수의 합산 값이 기준 값을 초과하는지 여부를 판별하는 단계; 및 상기 합산 값이 상기 기준 값을 초과할 경우, 해당 범죄 패턴을 위험 범죄 패턴으로 저장하는 단계;를 포함한다.According to an aspect of the present invention, there is provided a method of collecting crime data counted according to a predetermined classification criterion, the crime data indicating the number of crimes according to each item in the classification criterion; Calculating a crime risk index for each item on the basis of the crime data, wherein the crime risk index is a maximum value and a minimum value of the number of crimes of the item, the number of crimes of each item in the classification standard to which the item belongs, Lt; / RTI > Generating a crime pattern for a combination of items selected for each of the classification criteria based on the calculated crime risk index; Determining whether the sum of the crime risk indices of each item of the generated crime pattern exceeds a reference value; And storing the crime pattern as a crime pattern when the sum value exceeds the reference value.
본 발명에 따르면 범죄 발생 분류 기준에 따른 범죄 발생 확률을 위험 지수로 수치화 함으로써 범죄 발생 패턴을 파악하고, 예측할 수 있다.According to the present invention, crime occurrence patterns can be identified and predicted by quantifying the crime occurrence probability according to the crime occurrence classification standard as a risk index.
또한 본 발명의 실시예에 따르면 정확한 사용자의 위치 파악을 수행하고, 이를 통해 해당 사용자에게 최적화된 범죄 발생 패턴 정보를 제공할 수 있다. Also, according to the embodiment of the present invention, accurate location of the user can be grasped and crime occurrence pattern information optimized for the user can be provided.
도 1은 본 발명의 실시예에 따른 범죄 예측 시스템을 도시한 블록도를 나타낸다.1 is a block diagram illustrating a system for predicting a crime according to an embodiment of the present invention.
기 설정된 분류 기준 별로 계수된 범죄 데이터를 수집하는 단계, 상기 범죄 데이터는 해당 분류 기준에서의 각 항목에 따른 범죄 발생 횟수를 나타냄; Collecting the crime data counted according to a predetermined classification criterion, the crime data indicating the number of crimes according to each item in the classification criterion;
상기 범죄 데이터에 기초하여 상기 각 항목별 범죄 위험 지수를 산출하는 단계, 상기 범죄 위험 지수는 해당 항목의 범죄 발생 횟수, 해당 항목이 속한 분류 기준에서의 각 항목 별 범죄 발생 횟수의 최대값 및 최소값에 기 초하여 산출됨; Calculating a crime risk index for each item on the basis of the crime data, the crime risk index is calculated by multiplying the maximum value and the minimum value of the number of occurrences of the crime by each item in the category of the crime occurrence, Calculated based on;
상기 산출된 범죄 위험 지수에 기초하여, 상기 각 분류 기준 별로 선택된 항목의 조합에 대한 범죄 패턴을 생성하는 단계; Generating a crime pattern for a combination of items selected for each of the classification criteria based on the calculated crime risk index;
상기 생성된 범죄 패턴의 각 항목별 범죄 위험 지수의 합산 값이 기준 값을 초과하는지 여부를 판별하는 단계; 및 Determining whether the sum of the crime risk indices of each item of the generated crime pattern exceeds a reference value; And
상기 합산 값이 상기 기준 값을 초과할 경우, 해당 범죄 패턴을 위험 범죄 패턴으로 저장하는 단계; 를 포함하며,Storing the crime pattern as a dangerous crime pattern when the sum value exceeds the reference value; / RTI >
상기 범죄 위험 지수는 상기 최대값과 최소값 간의 차이에 대한 해당 항목의 범죄 발생 횟수의 비율에 기초하여 산출되며,The crime risk index is calculated on the basis of the ratio of the number of crime incidents of the item to the difference between the maximum value and the minimum value,
상기 범죄 패턴은 상기 각 분류 기준에서 상기 범죄 위험 지수가 높은 순서로 선택된 항목들의 조합이며,Wherein the crime pattern is a combination of items selected in descending order of the crime risk index in each of the classification criteria,
상기 범죄 데이터를 수집하는 단계는, 상기 분류 기준의 각 항목 별 범죄 발생 횟수의 형태로 표현된 정형 데이터를 수신하는 단계; 및 뉴스 데이터, SNS(Social Network Service) 데이터 및 웹 데이터 중 적어도 하나를 이용하여 범죄 관련 비정형 데이터를 획득하는 단계를 더 포함하고, The step of collecting the crime data may include receiving formatted data expressed in the form of the number of occurrences of crime for each item of the classification criteria; And acquiring crime related unstructured data using at least one of news data, social network service (SNS) data, and web data,
상기 정형 데이터 및 비정형 데이터를 함께 이용하여 상기 범죄 데이터를 수집하는 것을 특징으로 하는 범죄 패턴 분석을 이용한 범죄 예측 방법.And the criminal data is collected using the fixed data and the unstructured data together.
이하, 첨부된 도면을 참조하여 본 발명의 실시 예에 대하여 본 발명이 속하는 기술 분야에서 통상의 지식을 가진 자가 용이하게 실시할 수 있도록 상세히 설명한다. 그러나 본 발명은 여러 가지 상이한 형태로 구현될 수 있으며 이하에서 개시되는 실시 예에 한정되지 않는다. 또한 도면에서 본 발명을 명확하게 개시하기 위해서 본 발명과 관계없는 부분은 생략하였으며, 도면에서 동일하거나 유사한 부호들은 동일하거나 유사한 구성요소들을 나타낸다.Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Also, in order to clearly illustrate the present invention in the drawings, portions not related to the present invention are omitted, and the same or similar reference numerals denote the same or similar components.
본 발명의 목적 및 효과는 하기의 설명에 의해서 자연스럽게 이해되거나 보다 분명해질 수 있으며, 하기의 기재만으로 본 발명의 목적 및 효과가 제한되는 것은 아니다.The objects and effects of the present invention can be understood or clarified naturally by the following description, and the objects and effects of the present invention are not limited only by the following description.
본 발명의 목적, 특징 및 장점은 다음의 상세한 설명을 통하여 보다 분명해 질 것이다. 또한, 본 발명을 설명함에 있어서 본 발명과 관련된 공지 기술에 대한 구체적인 설명이, 본 발명의 요지를 불필요하게 흐릴 수 있다고 판단되는 경우에는 그 상세한 설명을 생략하기로 한다. 이하, 첨부된 도면을 참조하여 본 발명에 따른 실시예를 상세히 설명하기로 한다.The objects, features and advantages of the present invention will become more apparent from the following detailed description. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail. Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
도 1에 도시된 바와 같이 본 발명의 실시예에 따른 범죄 예측 시스템은 서버(100)와 사용자 단말(200)을 포함한다. 본 발명의 서버(100)는 범죄 데이터를 수집하고 범죄 발생 패턴을 생성 및 분석하여 사용자에게 제공하며, 통신 부(110), 데이터 수집부(120), 데이터 처리부(130), 범죄 패턴 생성부(140) 및 빅데이터 DB(150)를 포함할 수 있다.As shown in FIG. 1, a crime prediction system according to an embodiment of the present invention includes a server 100 and a user terminal 200. The server 100 of the present invention collects crime data and generates and analyzes a crime pattern to provide the data to a user. The server 100 includes a communication unit 110, a data collecting unit 120, a data processing unit 130, a crime pattern generating unit 140 and a big data DB 150. [
먼저, 통신부(110)는 외부 단말 또는 네트워크와 다양한 유선 또는 무선의 통신 방식을 사용하여 데이터를 송/ 수신할 수 있다. 이때, 사용 가능한 무선 통신 방식으로는 와이파이, LTE(Long Term Evolution), 블루투스, NFC(Near Field Communication), Zigbee, 적외선 통신 등이 있으며, 본 발명은 이에 한정하지 않는다.First, the communication unit 110 can transmit / receive data to / from an external terminal or a network using various wired or wireless communication systems. At this time, the available wireless communication methods include Wi-Fi, Long Term Evolution (LTE), Bluetooth, Near Field Communication (NFC), Zigbee, infrared communication, and the like.
다음으로, 데이터 수집부(120)는 상기 통신부(110)를 통해 범죄 관련 데이터를 수집한다. 본 발명의 실시예에서 범죄 데이터는 기 설정된 분류 기준 별로 계수된 범죄 발생 횟수를 나타낼 수 있다.Next, the data collecting unit 120 collects the crime-related data through the communication unit 110. In the embodiment of the present invention, the crime data may indicate the number of crime occurrences counted according to predetermined classification criteria.
본 발명의 실시예에 따르면, 데이터 수집부(120)는 범죄 관련 정형 데이터와 비정형 데이터를 함께 수집할 수 있다. 정형 데이터는 상기 분류 기준의 각 항목별 범죄 발생 횟수의 형태로 표현된 범죄 데이터를 가리키며, 대검찰청 등의 공공 기관으로부터 획득될 수 있다. 또한, 비정형 데이터는 뉴스 데이터, SNS(Social Network Service) 데이터 및 웹 데이터 중 적어도 하나를 이용하여 획득된다. 데이터 수집부(120)는 상기 뉴스 데이터 등에서 특정 범죄 사건에 대한 세부 정보를 추출할 수 있으며, 추출된 정보를 이용하여 범죄 데이터를 수집한다.According to the embodiment of the present invention, the data collection unit 120 may collect crime-related regular data and unstructured data together. Formal data refers to crime data expressed in the form of the number of occurrences of crime for each item in the classification standard, and can be obtained from public authorities such as the Supreme Prosecutor's Office. In addition, the atypical data is acquired using at least one of news data, social network service (SNS) data, and web data. The data collecting unit 120 may extract detailed information on a specific crime event from the news data or the like, and collects crime data using the extracted information.
또한,데이터수집부(120)는 사용자의 위치 정보 및 현재 시간 정보 등을 추가로 획득할 수 있다. 수집된 사용자의 위치 정보 및 현재 시간 정보 등은 데이터 처리부(130)로 전달된다.The data collection unit 120 may further acquire location information of the user, current time information, and the like. The collected location information of the user, current time information, and the like are transmitted to the data processing unit 130.
데이터 처리부(130)는 수집된 범죄 데이터에 기초하여 분류 기준의 각 항목별 범죄 위험 지수를 산출한다. 본 발명의 실시예에 따르면, 범죄 위험 지수는 해당 항목의 범죄 발생 횟수, 해당 항목이 속한 분류 기준에서의 각 항목 별 범죄 발생 횟수의 최대값 및 최소값에 기초하여 산출될 수 있다. 또한, 데이터 처리부(130)는 서버 (100) 내부의 데이터를 프로세싱하고, 프로세싱 된 데이터 또는 빅데이터 DB(150)에 저장된 데이터를 통신부(110)를 통해 사용자에게 제공할 수 있다.The data processing unit 130 calculates a crime risk index for each category of classification criteria based on the collected crime data. According to the embodiment of the present invention, the crime risk index can be calculated on the basis of the maximum value and the minimum value of the number of crimes of the item, and the number of crimes of each item in the classification standard to which the item belongs. The data processing unit 130 may process the data in the server 100 and provide the processed data or the data stored in the big data DB 150 to the user through the communication unit 110. [
범죄 패턴 생성부(140)는 산출된 범죄 위험 지수에 기초하여 범죄 패턴을 생성한다. 더욱 구체적으로 본 발명의 범죄 패턴은 각 분류 기준 별로 선택된 항목의 조합에 대한 정보를 나타내며, 상기 범죄 위험 지수에 기초하여 생성될 수 있다. 이에 대한 구체적인 설명은 후술하도록 한다.The crime pattern generator 140 generates a crime pattern based on the calculated crime risk index. More specifically, the crime pattern of the present invention represents information on a combination of items selected for each classification criterion, and can be generated based on the crime risk index. A detailed description thereof will be given later.
빅데이터 DB(140)는 다양한 디지털 데이터를 저장하며, 본 발명의 실시예에 따라 범죄 데이터 및 범죄 패턴 등의 정보를 저장할 수 있다. 일 실시예에 따르면, 빅데이터 DB(140)는 상기 범죄 패턴 생성부(140)에서 생성된 범죄 패턴들 중 각 항목별 범죄 위험 지수의 합산 값이 기준 값을 초과하는 패턴을 위험 범죄 패턴으로 지정하여 저장할 수 있다. 빅데이터 DB(140)는 하드디스크 드라이브, 플래시 메모리, RAM(Random Access Memory), SSD(Solid State Drive) 등의 다양한 디지털 데이터 저장 매체를 통해 구현될 수 있다.The big data DB 140 stores various digital data, and may store information such as crime data and crime patterns according to an embodiment of the present invention. According to one embodiment, the big data DB 140 may designate a pattern in which the sum of the crime risk indices for each item out of the crime patterns generated by the crime pattern generator 140 exceeds a reference value as a dangerous crime pattern . The big data DB 140 can be implemented through various digital data storage media such as a hard disk drive, a flash memory, a random access memory (RAM), and a solid state drive (SSD).
한편, 사용자 단말(200)은 서버(100)로부터 위험 범죄 패턴 정보를 수신하고, 이를 영상 및/또는 음성 정보로 출력할 수 있다. 본 발명에서 사용자 단말(200)은 다양한 형태의 디지털 디바이스, 이를테면 스마트폰, 데스크 탑, 랩탑, 태블릿 피씨, PDA(Personal Digital Assistant), 스마트 워치, HMD(Head Mounted Display) 등을 포함할 수 있다. 도 1에 도시된 바와 같이, 사용자 단말(200)은 통신부(210), 출력부(220) 및 제어부(230)를 포함할 수 있다. Meanwhile, the user terminal 200 may receive the dangerous crime pattern information from the server 100 and output it as video and / or audio information. The user terminal 200 may include various types of digital devices such as a smart phone, a desktop, a laptop, a tablet PC, a personal digital assistant (PDA), a smart watch, and a head mounted display (HMD). 1, the user terminal 200 may include a communication unit 210, an output unit 220, and a control unit 230.
통신부(210)는 서버 또는 네트워크와 다양한 유선 또는 무선의 통신 방식을 사용하여 데이터를 송/수신할 수 있다. 이때, 사용 가능한 무선 통신 방식으로는 와이파이, LTE(Long Term Evolution), 블루투스, NFC(Near Field Communication), Zigbee, 적외선 통신 등이 있음은 전술한 바와 같다.The communication unit 210 can transmit / receive data to / from the server or the network using various wired or wireless communication systems. At this time, the available wireless communication methods include Wi-Fi, Long Term Evolution (LTE), Bluetooth, Near Field Communication (NFC), Zigbee, and infrared communication.
출력부(220)는 다양한 형태의 영상 및/또는 음성 정보 출력 수단, 이를 테면 디스플레이 유닛, 스피커 등을 포함할 수 있다. 출력부(220)는 제어부(230)의 제어에 기초하여, 각종 정보를 영상 및/또는 음성 형태로 출력할 수 있다.The output unit 220 may include various types of video and / or audio information output means, such as a display unit, a speaker, and the like. The output unit 220 can output various kinds of information in the form of video and / or audio based on the control of the control unit 230. [
제어부(230)는 사용자 단말(200)의 각 부의 동작을 제어하며, 사용자 단말(200) 내부의 데이터를 프로세싱할 수 있다. 또한, 제어부(230)는 사용자 단말(200)의 각 부 간의 데이터 송수신을 제어할 수 있다. 사용자 단말 (200)의 제어부(230)는 서버(100)로부터 수신된 위험 범죄 패턴 정보를 출력부(220)가 통해 출력하도록 제어할 수 있다. 또는, 사용자 단말(200)의 제어부(230)는 상기 위험 범죄 패턴 정보에 기초한 경보 신호를 출력부(220)를 통해 출력할 수 있다.The control unit 230 may control the operation of each unit of the user terminal 200 and may process data in the user terminal 200. In addition, the control unit 230 can control data transmission / reception between the respective units of the user terminal 200. [ The control unit 230 of the user terminal 200 may control the output unit 220 to output the dangerous crime pattern information received from the server 100. [ Alternatively, the control unit 230 of the user terminal 200 may output an alarm signal based on the dangerous crime pattern information through the output unit 220.
본 발명의 실시예에 따르면, 사용자 단말(200)은 핑거프린트 기법을 이용하여 해당 단말의 위치를 측정할 수 있다. 더욱 구체적으로, 사용자 단말(200)은 복수의 AP(Access Point)로부터 수신된 통신 신호의 세기를 각각 측 정할 수 있으며, 측정된 신호의 세기에 기초하여 해당 단말의 위치 정보가 획득될 수 있다. 사용자 단말(200)은 획득된 해당 단말의 위치 정보를 서버(100)로 전송할 수 있다.According to the embodiment of the present invention, the user terminal 200 can measure the position of the corresponding terminal using the fingerprint technique. More specifically, the user terminal 200 can measure the intensity of a communication signal received from a plurality of access points (APs), and position information of the corresponding terminal can be obtained based on the intensity of the measured signal. The user terminal 200 may transmit the obtained location information of the corresponding terminal to the server 100.
범죄 발생 횟수는 다양한 분류 기준에 의해 분류될 수 있으며, 예를 들어 범죄 종류, 범죄 발생 지역, 범죄 발생 장소, 범죄 발생 시간 등에 따라 분류될 수 있다.The number of crime incidents can be classified according to various classification criteria, for example, according to type of crime, area of crime, location of crime, time of crime occurrence, and the like.
각 분류 기준은 적어도 하나의 항목으로 구성될 수 있다. 예를 들어, 범죄 종류(C)는 서로 다른 종류의 범죄 항목(C1, C2, C3, ...)으로 구성될 수 있는데, 이러한 범죄 항목은 대검찰청의 범죄 분류 기준에 따라 구성될 수도 있고, 본 발명의 시스템에서 고유하게 설정될 수도 있다. 범죄 데이터는 각 범죄 종류 항목(C1,C2,C3, ...)에 따른 범죄 발생 횟수에 대한 정보를 나타낸다.Each classification criterion can be composed of at least one item. For example, the type of crime (C) can be composed of different kinds of crime items (C1, C2, C3, ...), which can be organized according to crime classification criteria of the Supreme Prosecutors' And may be uniquely set in the system of the invention. The crime data shows information on the number of crimes according to each crime type item (C1, C2, C3, ...).
범죄 발생 지역(R)은 범죄 발생 지점의 지리적 구분에 따라 분류된 각 항목(R1,R2,R3,...)으로 구성될 수 있는데, 이를 테면 시,도 등의 광역 분류에 따라구 분되거나 더욱 세부적으로 각 구 별로 구분 될 수도 있다. 또한, 범죄 발생 장소(S)는 범죄 발생 지점의 유형에 따라 분류된 각 항목(S1, S2, S3, ...)으로 구성될 수 있는데, 이를 테면 아파트, 단독주택, 고속도로, 상점, 유흥 업소, 부대, 사무실, 학교, 창고, 지하철, 대중 교통 수단 등을 포함 할 수 있다.The crime occurrence area (R) can be composed of each item (R1, R2, R3, ...) classified according to the geographical division of the crime occurrence point, for example, More precisely, it may be divided into each phrase. In addition, the place of crime (S) can be composed of each item (S1, S2, S3, ...) classified according to the type of the crime spot. For example, apartment, single house, highway, , Units, offices, schools, warehouses, subways, public transportation, and the like.
범죄 발생 시간(T)의 각 항목(T1, T2, T3, ...)은 기 설정된 시간 단위로 구성되거나, 임의의 시간대 별로 새벽, 아침,낮,오전,오후,저녁,밤 등과 같이 구성될 수 있다. 또한, 실시예에 따라 범죄 분류 기준은 범죄 발생 요일(D)을 포함할 수 있는데, 범죄 발생 요일(D)의 각 항목(D1, D2, D3, ...)은 한 주의 각 요일 즉, 일요일, 월요일, ... , 토요일을 나타낼 수 있다.Each item (T1, T2, T3, ...) of the crime occurrence time (T) may be configured as a predetermined time unit, or may be configured as an arbitrary time zone such as dawn, morning, day, morning, afternoon, evening, . In addition, according to the embodiment, the crime classification standard may include a day of crime occurrence (D). Each item (D1, D2, D3, ...) of the day of crime occurrence (D) , Monday, ..., Saturday.
즉, 범죄 패턴 생성부(140)는 각 범죄 종류(C)에 따라 범죄 발생 지역(R) 중 범죄 위험 지수(INDRn)가 높은 항목(Rn), 범죄 발생 시간(T) 중 범죄 위험 지수(INDTn)가 높은 항목(Tn), 범죄 발생 요일(D) 중 범죄 위험 지수 (INDDn)가 높은 항목(Dn), 범죄 발생 장소(S) 중 범죄 위험 지수(INDSn)가 높은 항목(Sn)을 선택하여 범죄 패턴 P(R, T, D, S)를 생성할 수 있다.In other words, the crime pattern generator 140 calculates the crime index INDn of the crime occurrence area R according to the crime type C, the item Rn having the high INDRn value, (Dn) of high crime index (INDDn) and high crime index (INDSn) (Sn) among high crime index (Tn), crime day It is possible to generate the crime pattern P (R, T, D, S).
본 발명의 서버(100)는 상기 생성된 범죄 위험 지수 합산 값 INDRTDS가 기준 값을 초과하는지 여부를 판별하고, 기준 값을 초과할 경우 해당 범죄 패턴을 위험 범죄 패턴으로 빅데이터 DB(150)에 저장한다. 이에 따라, 빅데이터DB(150)는 범죄 위험 지수의 합산 값이 기준값보다 큰 범죄 패턴들의 정보를 저장할 수 있다.The server 100 of the present invention determines whether or not the generated crime risk index sum value INDRTDS exceeds the reference value and if the crime index sum value INCDTDS exceeds the reference value, the server 100 stores the crime pattern as a risky crime pattern in the big data DB 150 do. Accordingly, the big data DB 150 can store information on crime patterns in which the sum of the crime risk indices is larger than the reference value.
한편, 범죄 패턴 생성부(140)는 각 분류 기준별 항목을 변경해가며 또 다른 범죄 패턴들을 생성하고, 해당 범죄 패턴의 각 항목별 범죄 위험 지수의 합산값을산출할 수 있다. 범죄패턴생성부(140)는 범죄 패턴 생성 시 해당 패턴이 기존에 생성된 범죄 패턴인지 여부를 확인하여, 중복된 범죄 패턴이 생성되지 않도록 한다. 이때, 서로 다른 범죄 패턴 생성을 위해, 각 분류 기준에서 범죄 위험 지수가 높은 순서로 항목들이 선택될 수 있다. 해당 범죄 패턴의 각 항목별 범죄 위험 지수의 합산 값이 기준 값을 초과하면, 해당 범죄 패턴은 위험 범죄 패턴으로 빅데이터 DB(150)에 저장된다. 본 발명의 서버(100)는 DB(150)에 저장된 범죄 패턴 정보 특히, 위험 범 죄패턴정보등에기초하여사용자에게범죄예측정보를제공할수있다.Meanwhile, the crime pattern generator 140 may generate another crime pattern by changing items for each classification criterion, and calculate the sum of crime risk indexes for each item of the crime pattern. The crime pattern generation unit 140 checks whether the corresponding pattern is a crime pattern generated at the time of generating the crime pattern, thereby preventing duplicate crime patterns from being generated. In order to generate different crime patterns, items can be selected in descending order of crime risk index in each classification criterion. If the sum of the crime risk indices for each item of the crime pattern exceeds the reference value, the crime pattern is stored in the big data DB 150 as a dangerous crime pattern. The server 100 of the present invention can provide crime prediction information to a user based on crime pattern information stored in the DB 150, in particular, dangerous crime pattern information and the like.
사용자 단말(200)은 복수의 AP(Access Point, 50)로부터 수신된 신호의 세기를 측정하여 해당 단말의 위치를 추정 할 수 있다. 사용자 단말(200)은 AP 50a, 50b, 50c로부터 프로브 응답(Probe Response) 신 호 또는 그 밖의 다양한 신호를 수신할 수 있으며, 수신된 AP 신호에는 해당 AP의 위치 정보가 포함될 수 있다. 사용자 단말(200)은 수신된 복수의 AP 신호 각각의 위치 정보, 수신된 신호의 세기 정보 등에 기초하여 삼각 측 량법을 이용하여 해당 디바이스의 위치 정보를 획득할 수 있다.The user terminal 200 can estimate the position of a corresponding terminal by measuring the strength of a signal received from a plurality of APs (Access Points) 50. The user terminal 200 may receive a probe response signal or various signals from the APs 50a, 50b, and 50c, and the received AP signal may include location information of the corresponding AP. The user terminal 200 can acquire position information of the corresponding device using the triangulation method based on the position information of each of the plurality of received AP signals, the intensity information of the received signal, and the like.
사용자의 위치를 추정하고자 하는 공간을 nxm격자 형태로 설정하고 각 지점P(1,1),P(1, 2), ... , P(n, m)에서의 수신되는 복수의 AP(AP1, AP2, ... , APk) 신호의 세기를 측정한다. 측정 오차를 줄이기 위해 각 지점에서 복수 회에 걸쳐 AP신호를 측정하고, 측정된 값의 평균 값이 DB에 저장될 수 있다. 이와 같이 DB가 구축되면, 사용자 단말(200)에서 수신되는 복수의 AP 신호의 세기의 조합에 대응하는 해당 단말의 위치 정 보를 제공할 수 있게 된다. 즉, 임의의 위치 P(x, y)에서 사용자 단말(200)은 복수의 AP로부터 수신되는 신호의 세기를 측정하고, 측정된 신호의 세기를 DB에 저장된 각 지점에서의 기 측정 값과 비교하여, 수신 신호 세기 의오차가가장적은지점을해당단말의위치로결정할수있다.A space for estimating the location of the user is set in the form of nxm grid and a plurality of APs AP1 (1, 2), ..., P (n, m) , AP2, ..., APk) are measured. To reduce the measurement error, the AP signal is measured multiple times at each point, and the average value of the measured values can be stored in the DB. When the DB is constructed as described above, it is possible to provide location information of a corresponding terminal corresponding to a combination of strengths of a plurality of AP signals received at the user terminal 200. [ That is, at an arbitrary position P (x, y), the user terminal 200 measures the intensity of a signal received from a plurality of APs, compares the intensity of the measured signal with a pre-measured value at each point stored in the DB , It is possible to determine the position where the error of the received signal strength is the smallest as the position of the corresponding terminal.
사용자 단말(200)은 AP 신호뿐만 아니라 GPS, 블루투스 비콘 신호 또는 이들의 조합을 이용하여 사용자의 위치 정보를 획득할 수도 있다. 이와 같이 사용자 단말(200)의 위치 정보가 획득되면, 해당 단말의 위치 정보는 서버(100)로 전송된다. 위치 정보를 수신한 서버(100)는 빅데이터 DB(150)에 저장된 위험 범죄패턴들중해당사용자단말(200)의위치정보에대응하는위험범죄패턴을검색할수있다. 즉,서버 (100)는 해당 사용자 단말(200)의 위치 정보에 매칭되는 범죄 발생 지역/범죄 발생 장소 항목을 갖는 위험 범죄 패턴을 검색한다.The user terminal 200 may acquire the location information of the user using the GPS signal, the Bluetooth beacon signal, or a combination thereof as well as the AP signal. When the location information of the user terminal 200 is obtained, the location information of the terminal is transmitted to the server 100. The server 100 receiving the location information can search for a dangerous crime pattern corresponding to the location information of the corresponding user terminal 200 among the dangerous crime patterns stored in the big data DB 150. That is, the server 100 searches for a dangerous crime pattern having a crime occurrence area / crime occurrence place item matched with the position information of the corresponding user terminal 200.
한편, 본 발명의 서버(100)는 현재 시간 정보를 더 획득할 수 있다. 서버(100)는 GPS 모듈을 탑재하여 현재 시간 정보를 수신할 수 있으며, 다른 실시예에 따르면 사용자 단말(200)로부터 해당 단말의 위치 정보와 함께 현재 시간 정보를 수신할 수 있다. 현재 시간 정보를 수신한 서버(100)는 빅데이터 DB(150)에 저장된 위험 범죄 패턴들 중 현재 시간에 대응하는 범죄 발생 시간/범죄 발생 요일 항목을 갖는 위험 범죄 패턴을 검색할 수 있다. 따라서, 서버(100)가 사용자 단말(200)의 위치 정보와 현재 시간 정보를 함께 획득할 경우, 해당 위치 정보 및 현재 시간 정보에 매칭되는 범죄 발생 지역/범죄 발생 장소/범죄 발생 시간/범죄 발생 요일 항목을 갖 는위험범죄패턴이검색될수있다.Meanwhile, the server 100 of the present invention can obtain current time information. The server 100 may receive the current time information by mounting the GPS module and may receive the current time information together with the location information of the terminal from the user terminal 200 according to another embodiment. The server 100 receiving the current time information can search for a dangerous crime pattern having the crime occurrence time / crime occurrence day item corresponding to the current time among the dangerous crime patterns stored in the big data DB 150. [ Accordingly, when the server 100 acquires the location information of the user terminal 200 together with the current time information, the location information of the crime occurrence / crime occurrence place / crime occurrence time / crime occurrence day Risk crime patterns with items can be retrieved.
서버(100)는 이와 같이 검색된 위험 범죄 패턴 정보를 해당 사용자 단말(200)에 전송할 수 있다. 서버(100)로 부터 위험 범죄 패턴 정보를 수신한 사용자 단말(200)은 출력부(220)를 통해 영상 및/또는 음성으로 해당 정보 를 출력한다.The server 100 may transmit the detected dangerous crime pattern information to the corresponding user terminal 200. [ The user terminal 200 receiving the dangerous crime pattern information from the server 100 outputs the information through the output unit 220 with the image and / or voice.
서버는 기 설정된 분류 기준 별로 계수된 범죄 데이터를 수집한다(S110). 상기 범죄 데이터는 해당 분류 기준에서의각항목에따른범죄발생횟수를나타낸다. 본발명의실시예에따르면,상기분류기준은범죄 종류, 범죄 발생 지역, 범죄 발생 장소, 범죄 발생 시간, 범죄 발생 요일 등을 포함할 수 있다.The server collects the crime data counted according to predetermined classification criteria (S110). The above-mentioned crime data represents the number of crimes occurring according to each item in the classification standard. According to an embodiment of the present invention, the classification criteria may include a crime type, a crime occurrence area, a crime occurrence place, a crime occurrence time, a crime occurrence day, and the like.
다음으로, 서버는 범죄 데이터에 기초하여 각 항목 별 범죄 위험 지수를 산출한다(S120). 상기 범죄 위험 지수 는해당항목의범죄발생횟수에비례할수있다. 본발명의실시예에따르면,상기범죄위험지수는해당 항목의 범죄 발생 횟수뿐만 아니라 해당 항목이 속한 분류 기준에서의 각 항목 별 범죄 발생 횟수의 최대값 및 최소값에기초하여산출될수있다. 즉, 범죄 위험 지수는 상기 최대값과 최소값 간의 차이에대한 해당 항목의 범죄 발생 횟수의 비율에 기초하여 산출된다.Next, the server calculates the crime risk index for each item based on the crime data (S120). The crime risk index may be proportional to the number of crimes in the item. According to the embodiment of the present invention, the crime risk index can be calculated based on the maximum value and the minimum value of the number of occurrences of crime for each item in the classification standard to which the item belongs as well as the number of crime occurrence of the item. That is, the crime risk index is calculated based on the ratio of the number of crimes of the item to the difference between the maximum value and the minimum value.
다음으로, 서버는 산출된 범죄 위험 지수에 기초하여 각 분류 기준 별로 선택된 항목의 조합에 대한 범죄 패턴을 생성한다(S130). 상기 범죄 패턴은 각 분류 기준에서 범죄 위험 지수가 높은 순서로 선택된 항목들의 조합이다.Next, the server generates a crime pattern for a combination of items selected for each classification criterion based on the calculated crime risk index (S130). The crime pattern is a combination of items selected in descending order of crime risk index in each classification standard.
다음으로, 서버는 생성된 범죄 패턴의 각 항목별 범죄 위험 지수의 합산 값이 기준값을 초과하는지 여부를판별한다(S140). 만약, 범죄 위험 지수의 합산 값이 기준값을 초과하면, 서버는 해당 범죄 패턴을 위험 범죄 패턴으로 빅데이터 DB에 저장한다(S150). 서버는 다시 S130 단계로 돌아가서, 각 분류 기준별 항목을 변경해가며 또 다른 범죄 패턴들을 생성하고, 해당 범죄 패턴에 대하여 S140단계 및 S150단계를 반복한다. 이때, 서버는 해당 패턴이 기존에 생성된 범죄 패턴과 동일한지 여부를 확인하여, 중복된 범죄 패턴이 생성되지 않도록 한다.Next, the server determines whether the sum of the crime risk indices of each item of the generated crime pattern exceeds the reference value (S140). If the sum of the crime risk index exceeds the reference value, the server stores the crime pattern in the big data DB as a crime pattern (S150). The server returns to step S130 and changes the item of each classification criterion to generate another crime pattern and repeats steps S140 and S150 for the crime pattern. At this time, the server confirms whether or not the pattern is the same as the previously generated crime pattern, thereby preventing duplicate crime patterns from being generated.
상기한 본 발명의 바람직한 실시 예는 예시의 목적으로 개시된 것이고, 본 발명에 대해 통상의 지식을 가진 당업자라면 본 발명의 사상과 범위 안에서 다양한 수정, 변경 및 부가가 가능할 것이며, 이러한 수정, 변경 및 부가는 상기의 특허청구 범위에 속하는 것으로 보아야 할 것이다. It will be apparent to those skilled in the art that various modifications, additions and substitutions are possible, without departing from the spirit and scope of the invention as defined by the appended claims. Should be regarded as belonging to the above-mentioned patent claims.
본 발명이 속하는 기술분야에서 통상의 지식을 가진 자라면, 본 발명의 기술적 사상을 벗어나지 않는 범위 내에서, 여러 가지 치환, 변형 및 변경이 가능하므로, 본 발명은 전술한 실시 예 및 첨부된 도면에 의해 한정되는 것이 아니다.It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the inventive concept as defined by the appended claims. But is not limited thereto.
상술한 예시적인 시스템에서, 방법들은 일련의 단계 또는 블록으로써 순서도를 기초로 설명되고 있지만, 본 발명은 단계들의 순서에 한정되는 것은 아니며, 어떤 단계는 상술한 바와 다른 단계와 다른 순서로 또는 동시에 발생할 수 있다. 또한, 당업자라면 순서도에 나타낸 단계들이 배타적이지 않고, 다른 단계가 포함되거나 순서도의 하나 또는 그 이상의 단계가 본 발명의 범위에 영향을 미치지 않고 삭제될 수 있음을 이해할 수 있을 것이다.In the above-described exemplary system, the methods are described on the basis of a flowchart as a series of steps or blocks, but the present invention is not limited to the order of the steps, and some steps may occur in different orders . It will also be understood by those skilled in the art that the steps shown in the flowchart are not exclusive and that other steps may be included or that one or more steps in the flowchart may be deleted without affecting the scope of the invention.
본 발명에 따르면 범죄 발생 분류 기준에 따른 범죄 발생 확률을 위험 지수로 수치화 함으로써 범죄 발생 패턴을 파악하고, 예측할 수 있다.According to the present invention, crime occurrence patterns can be identified and predicted by quantifying the crime occurrence probability according to the crime occurrence classification standard as a risk index.
또한 본 발명의 실시예에 따르면 정확한 사용자의 위치 파악을 수행하고, 이를 통해 해당 사용자에게 최적화된 범죄 발생 패턴 정보를 제공할 수 있다.Also, according to the embodiment of the present invention, accurate location of the user can be grasped and crime occurrence pattern information optimized for the user can be provided.

Claims (1)

  1. 기 설정된 분류 기준 별로 계수된 범죄 데이터를 수집하는 단계, 상기 범죄 데이터는 해당 분류 기준에서의 각 항목에 따른 범죄 발생 횟수를 나타냄; Collecting the crime data counted according to a predetermined classification criterion, the crime data indicating the number of crimes according to each item in the classification criterion;
    상기 범죄 데이터에 기초하여 상기 각 항목별 범죄 위험 지수를 산출하는 단계, 상기 범죄 위험 지수는 해당 항 목의 범죄 발생 횟수, 해당 항목이 속한 분류 기준에서의 각 항목 별 범죄 발생 횟수의 최대값 및 최소값에 기 초하여 산출됨; Calculating a crime risk index for each item on the basis of the crime data, wherein the crime risk index is a maximum value and a minimum value of the number of crimes occurring in each item in the classification category to which the item belongs, Calculated based on;
    상기 산출된 범죄 위험 지수에 기초하여, 상기 각 분류 기준 별로 선택된 항목의 조합에 대한 범죄 패턴을 생성하는 단계; Generating a crime pattern for a combination of items selected for each of the classification criteria based on the calculated crime risk index;
    상기 생성된 범죄 패턴의 각 항목별 범죄 위험 지수의 합산 값이 기준 값을 초과하는지 여부를 판별하는 단계; 및 Determining whether the sum of the crime risk indices of each item of the generated crime pattern exceeds a reference value; And
    상기 합산 값이 상기 기준 값을 초과할 경우, 해당 범죄 패턴을 위험 범죄 패턴으로 저장하는 단계; 를 포함하며,Storing the crime pattern as a dangerous crime pattern when the sum value exceeds the reference value; / RTI >
    상기 범죄 위험 지수는 상기 최대값과 최소값 간의 차이에 대한 해당 항목의 범죄 발생 횟수의 비율에 기초하여 산출되며,The crime risk index is calculated on the basis of the ratio of the number of crime incidents of the item to the difference between the maximum value and the minimum value,
    상기 범죄 패턴은 상기 각 분류 기준에서 상기 범죄 위험 지수가 높은 순서로 선택된 항목들의 조합이며,Wherein the crime pattern is a combination of items selected in descending order of the crime risk index in each of the classification criteria,
    상기 범죄 데이터를 수집하는 단계는, 상기 분류 기준의 각 항목 별 범죄 발생 횟수의 형태로 표현된 정형 데이터를 수신하는 단계; 및 뉴스 데이터, SNS(Social Network Service) 데이터 및 웹 데이터 중 적어도 하나를 이용하여 범죄 관련 비정형 데이터를 획득하는 단계를 더 포함하고, The step of collecting the crime data may include receiving formatted data expressed in the form of the number of occurrences of crime for each item of the classification criteria; And acquiring crime related unstructured data using at least one of news data, social network service (SNS) data, and web data,
    상기 정형 데이터 및 비정형 데이터를 함께 이용하여 상기 범죄 데이터를 수집하는 것을 특징으로 하는 범죄 패턴 분석을 이용한 범죄 예측 방법.And the criminal data is collected using the fixed data and the unstructured data together.
PCT/KR2017/013563 2017-11-24 2017-11-24 System for predicting location-specific crime occurrence probability on basis of machine learning algorithm by using crime-related big data WO2019103205A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/KR2017/013563 WO2019103205A1 (en) 2017-11-24 2017-11-24 System for predicting location-specific crime occurrence probability on basis of machine learning algorithm by using crime-related big data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/KR2017/013563 WO2019103205A1 (en) 2017-11-24 2017-11-24 System for predicting location-specific crime occurrence probability on basis of machine learning algorithm by using crime-related big data

Publications (1)

Publication Number Publication Date
WO2019103205A1 true WO2019103205A1 (en) 2019-05-31

Family

ID=66631592

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2017/013563 WO2019103205A1 (en) 2017-11-24 2017-11-24 System for predicting location-specific crime occurrence probability on basis of machine learning algorithm by using crime-related big data

Country Status (1)

Country Link
WO (1) WO2019103205A1 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110175790A (en) * 2019-06-04 2019-08-27 广东铭太信息科技有限公司 A method of it is counted for multiple crime case
CN116028881A (en) * 2023-03-15 2023-04-28 深圳市大数据研究院 Prediction method and system for reappearance limiting behavior based on multidimensional database

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20120051246A (en) * 2010-11-12 2012-05-22 (주)에스웨어 Crime index interlocking type system for controlling emergency situation using cellular phone
KR20140065568A (en) * 2012-11-16 2014-05-30 목원대학교 산학협력단 Crime map server using active data collection and crime risk area notification service methodes
KR20160104223A (en) * 2015-02-26 2016-09-05 광운대학교 산학협력단 A Method for a Crime Prediction Using a Crime Pattern Analysis based on Big Data and A System for the Same
KR20160104221A (en) * 2015-02-26 2016-09-05 광운대학교 산학협력단 A Method for Estimating a Crime Risk Using Big Data and A System for Reporting the Crime Risk Using thereof
KR20160105151A (en) * 2015-02-27 2016-09-06 광운대학교 산학협력단 A Method for a Crime Prediction by Estimating a Cell Value and A System for the Same

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20120051246A (en) * 2010-11-12 2012-05-22 (주)에스웨어 Crime index interlocking type system for controlling emergency situation using cellular phone
KR20140065568A (en) * 2012-11-16 2014-05-30 목원대학교 산학협력단 Crime map server using active data collection and crime risk area notification service methodes
KR20160104223A (en) * 2015-02-26 2016-09-05 광운대학교 산학협력단 A Method for a Crime Prediction Using a Crime Pattern Analysis based on Big Data and A System for the Same
KR20160104221A (en) * 2015-02-26 2016-09-05 광운대학교 산학협력단 A Method for Estimating a Crime Risk Using Big Data and A System for Reporting the Crime Risk Using thereof
KR20160105151A (en) * 2015-02-27 2016-09-06 광운대학교 산학협력단 A Method for a Crime Prediction by Estimating a Cell Value and A System for the Same

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110175790A (en) * 2019-06-04 2019-08-27 广东铭太信息科技有限公司 A method of it is counted for multiple crime case
CN116028881A (en) * 2023-03-15 2023-04-28 深圳市大数据研究院 Prediction method and system for reappearance limiting behavior based on multidimensional database
CN116028881B (en) * 2023-03-15 2023-06-02 深圳市大数据研究院 Prediction method and system for reappearance limiting behavior based on multidimensional database

Similar Documents

Publication Publication Date Title
EP2122518B1 (en) Method for equitable placement of a limited number of sensors for wide area surveillance
KR20160104223A (en) A Method for a Crime Prediction Using a Crime Pattern Analysis based on Big Data and A System for the Same
CN109831742B (en) Monitoring method and system based on terminal detection
KR101687477B1 (en) A Method for Providing Event Occurrence Information Using Big Data and A System for the Same
KR101445045B1 (en) System and method for emergency alerting using the sameusing integrated management service
KR20170112904A (en) Risk early warning method and device
WO2019103205A1 (en) System for predicting location-specific crime occurrence probability on basis of machine learning algorithm by using crime-related big data
KR20180118979A (en) Method and apparatus for risk detection, prediction, and its correspondence for public safety based on multiple complex information
Wang et al. A Modified Inverse Distance Weighting Method for Interpolation in Open Public Places Based on Wi‐Fi Probe Data
CN106066707A (en) The physical separation of detection portable set
KR101738745B1 (en) A Method for a Crime Prediction by Estimating a Cell Value and A System for the Same
KR101660160B1 (en) A Method for Estimating a Crime Risk Using Big Data and A System for Reporting the Crime Risk Using thereof
JP2018197679A (en) Earthquake alarm system
CN112953952A (en) Industrial security situation awareness method, platform, electronic device and storage medium
KR20200102390A (en) Mobile urgent disaster reporting and processing system
CN112308325A (en) Thermodynamic diagram generation method and device
KR20180130806A (en) Crime-Occurrence Prediction System by Location Based on Crime-related Big Data
CN110139278B (en) Method of safety type collusion attack defense system under Internet of vehicles
Meng et al. A dynamic emergency decision support model for emergencies in urban areas
CN114926795B (en) Method, device, equipment and medium for determining information relevance
CN103929322A (en) Communication Monitor And Prediction Method
CN205449948U (en) Visual system of data that gaseous benzene series physical prospecting was surveyed
Sinnott et al. Urban traffic analysis using social media data on the cloud
CN110796044A (en) Target area security monitoring method and device
Mostafa Abdelkhalek et al. Dynamic User-Centric Clustered Workplaces for COVID-19 Control Measures Based on Geofencing and Deep Learning

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17932672

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17932672

Country of ref document: EP

Kind code of ref document: A1