WO2023197350A1 - 一种基于wifi匹配的传染病易感人群接触追踪预筛选方法 - Google Patents

一种基于wifi匹配的传染病易感人群接触追踪预筛选方法 Download PDF

Info

Publication number
WO2023197350A1
WO2023197350A1 PCT/CN2022/088158 CN2022088158W WO2023197350A1 WO 2023197350 A1 WO2023197350 A1 WO 2023197350A1 CN 2022088158 W CN2022088158 W CN 2022088158W WO 2023197350 A1 WO2023197350 A1 WO 2023197350A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
wifi
record
data
dangerous
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/CN2022/088158
Other languages
English (en)
French (fr)
Inventor
贺诗波
李超
余睿伟
吴旻诚
陈积明
程鹏
刘宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang University
Original Assignee
Zhejiang University
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 Zhejiang University filed Critical Zhejiang University
Priority to JP2022555088A priority Critical patent/JP7564565B2/ja
Priority to US17/925,317 priority patent/US12087451B2/en
Publication of WO2023197350A1 publication Critical patent/WO2023197350A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/80—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/22—Indexing; Data structures therefor; Storage structures
    • G06F16/2228—Indexing structures
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/22—Indexing; Data structures therefor; Storage structures
    • G06F16/2282—Tablespace storage structures; Management thereof
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/22—Indexing; Data structures therefor; Storage structures
    • G06F16/2291—User-Defined Types; Storage management thereof
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
    • H04W12/02—Protecting privacy or anonymity, e.g. protecting personally identifiable information [PII]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W8/00—Network data management
    • H04W8/18—Processing of user or subscriber data, e.g. subscribed services, user preferences or user profiles; Transfer of user or subscriber data
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W8/00—Network data management
    • H04W8/26—Network addressing or numbering for mobility support
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Definitions

  • the present invention relates to the detection of susceptible populations to infectious diseases, and in particular to a contact tracing pre-screening method for susceptible populations to infectious diseases based on WiFi matching.
  • the novel coronavirus epidemic that broke out in early 2020 has had a huge impact on production and life around the world.
  • March 2022 the cumulative number of confirmed cases worldwide has exceeded 470 million, and the cumulative death toll has exceeded 6 million.
  • the World Health Organization has declared the 2019 coronavirus
  • the viral disease (COVID-19) has been identified as one of the most serious public health crises in history.
  • the COVID-19 epidemic has brought huge losses of life and property to the world, caused serious divisions among countries, societies, and people, and profoundly changed the world structure.
  • contact tracing In the process of epidemic prevention and control, contact tracing is a very critical step. In public health, contact tracing refers to the process of identifying contacts who may have come into contact with an infected person and subsequently collecting further information about those contacts. Since many infectious diseases, including COVID-19, need to be spread among people through respiratory tract transmission, contact transmission, etc., that is, person-to-person contact, it is important to detect these high-risk individuals who have been in contact with infected people as soon as possible Crowds play a very important role in promptly isolating potentially infected persons.
  • Absolute location data Use absolute location information, such as GPS location, location relative to static WiFi hotspots or communication base stations, etc. These data can usually be expressed in the form of geographical coordinates. These data may not be accurate enough for contact tracing of infectious diseases, but through large amounts of such data, the movement patterns of specific users can be discovered and clustered.
  • GPS data has limited accuracy indoors, and involves sensitive information such as a person’s specific location and trajectory; Bluetooth point-to-point contact data requires the large-scale use of professional equipment, which is expensive and difficult to popularize. .
  • the applicant proposes a contact tracing pre-screening method for infectious disease susceptible groups based on WiFi matching.
  • the purpose of the present invention is to make up for the shortcomings of existing application technology and provide a more comprehensive contact tracing method for susceptible groups to infectious diseases by using different types of data.
  • This method can use the WiFi access records of historical patients diagnosed with infectious diseases to construct dangerous
  • the WiFi database provides the user's infection risk after obtaining daily updated user WiFi access records, improving the efficiency of finding susceptible groups.
  • a contact tracing pre-screening method for infectious disease susceptible groups based on WiFi matching including the following steps:
  • User anonymous identity code record matching Obtain the confirmed user's unique anonymous identity code data table through the database, use hive sql to search for the WiFi connection records corresponding to these IDs in the total database, create data table partitions according to date, and save them as original data records.
  • Construct a dangerous WiFi database use the information data set after code mapping in step (2)
  • the dangerous WiFi database as dict dan and extract it Get the WiFi connection record of the confirmed user and get the WiFi code in the record Put it into dict dan , and record the WiFi income.
  • the date in the dangerous WiFi database is the current date day now ; if the WiFi already exists, update its income date to day now instead, and then change the income date Records whose difference from the current date day now exceeds the dangerous WiFi disappearance threshold thr dan will be deleted;
  • step 1) in the matching of user anonymous identity code records, the hive sql tool is used to match the unique anonymous identity code of the confirmed user with the national user WiFi connection record database, and generate a partitioned confirmed user data record table for subsequent processing , the specific implementation method is as follows:
  • the original database contains two parts: 1.1.1) Confirmed user information data table, including confirmed user anonymous identity code gid_confirm, diagnosis date confirm_day; 1.1.2) National user WiFi connection record data table, including user anonymous identity code gid, connection WiFi mac address mac, record generation date day, record generation area, record generation time hour, record time times of WiFi connections, use the same user anonymous identity code of the two databases as an index, use hive sql tool to generate confirmed users WiFi connection record data table;
  • the national user WiFi connection record data table is partitioned by area and date.
  • it is necessary to determine the target screening area (its area code is AREACORE) and the target inspection area in advance. Starting and ending time (BEGINDAY and ENDDAY respectively);
  • a two-way mapping dictionary is used to compress and restore the original user's anonymous identity code and the WiFi mac address.
  • the specific implementation method is as follows:
  • a two-way mapping dictionary class TwoWayDict Construct a two-way mapping dictionary class TwoWayDict. Its characteristic is that after inputting the original uncompressed string, a two-way mapping from uncompressed string to compressed string and compressed string to uncompressed string is formed inside the dictionary, that is, the uncompressed string and Each compressed string is saved as a key and corresponding value to save storage space and memory in subsequent processing. When the prediction result is obtained, the original uncompressed string is also obtained based on the two-way dictionary;
  • step 3 constructs a dangerous WiFi database, adds the mac address in the historical user WiFi connection data to the dangerous WiFi database according to the date, and dynamically deletes expired WiFi to reduce the false positive rate.
  • the specific implementation method is as follows:
  • dict dan uses the WiFi mac address as the key and the diagnosis date as the value. If there are duplicate mac addresses, select the one with the latest date and save it as the value;
  • step 4) of the infection risk judgment the daily updated user WiFi connection data is compared with the dangerous WiFi database to obtain the coincidence rate, and whether the user is a dangerous user is judged based on the threshold.
  • the coincidence rate r user of user user is calculated as:
  • W user is the WiFi the user connected to that day
  • W dan is the dangerous WiFi database
  • represents the number of elements in the set S
  • R user is higher than the risk judgment threshold thr user , it will be judged as a dangerous user, and key screening should be carried out on it.
  • the present invention is a contact tracing pre-screening method for infectious disease susceptible groups based on WiFi matching. It mainly uses user WiFi connection records collected based on mobile devices and anonymous identity code lists of confirmed users. Through record matching, information compression, and construction of a dangerous WiFi database, Calculate the dangerous WiFi coincidence rate of users and other steps to obtain the final list of dangerous users.
  • This invention analyzes, corrects, and reduces the dimensionality of WiFi connection record data. Compared with contact tracing methods, it can save computing resources, reduce processing time, and improve screening efficiency without involving private information such as location or using additional equipment. , providing another dimension of more comprehensive data support for the subsequent search for high-risk and susceptible groups that may have contact with confirmed users.
  • Figure 1 is a flow chart of the digital contact tracing pre-screening method based on WiFi matching according to the present invention
  • Figure 2 is a diagram showing the comparison results of precision rate and recall rate of screening among user groups according to the embodiment of the present invention.
  • This embodiment collects user WiFi connection record data collected from a certain place from January 1, 2020 to March 20, 2020, as well as anonymous information of users diagnosed with a certain respiratory infectious disease within this time period.
  • the specific variables and related data information included in the data set are shown in Table 1, Table 2, Table 3, and Table 4:
  • Table 1 WiFi connection record data of some users in a certain place
  • Field name type of data illustrate gid string User anonymous identity code hour int The time period during which records are generated mac string The mac address of the WiFi connection times int Record the number of times the WiFi is connected during the generation period day int The date the record was generated area int Record the area code of the region you belong to
  • Field name type of data illustrate confirm_gid string Confirmed user anonymous identity code confirm_day int The user’s diagnosis date day int The date this record belongs to
  • the default implementation data set of the contact tracing pre-screening method for susceptible groups to infectious diseases is the above-mentioned WiFi connection data of users in a certain place and data of users diagnosed with a certain infectious disease.
  • the result of the method is the obtained list of dangerous users, and its detailed implementation steps details as follows:
  • Match WiFi connection records based on the user’s anonymous identity code Use the hive sql tool to extract sub-tables from the national user WiFi connection record table based on the record date and the selected region code to improve matching efficiency. Use the same user anonymous identity code as the filter. Conditions, retain the user anonymous identity code, WiFi connection mac address, record generation date, and user diagnosis date fields in the corresponding records, and use the record generation date as the partition index to partition the generated confirmed user WiFi connection record data table. In order to verify the prediction effect, in this embodiment, some records of healthy users are sampled and processed in the same way, and the diagnosis date field of healthy users is processed by default. At the same time, add a seed field to the table. Seed equal to 0 indicates healthy users, and seed equal to 1 indicates confirmed users.
  • thr user is set to 0.1, and the dangerous user set of the day is constructed in the following form:
  • step 4) Map the current compressed user ID back to the original user anonymous identity code, and conduct key screening of these users.
  • the user code mapping result in step 4) is:
  • the daily precision rate and recall rate are shown in Figure 1. Considering that the recall rate index is more important than the accuracy rate in infectious disease screening, the present invention can better realize the screening of dangerous persons.
  • the present invention is a contact tracing pre-screening method for infectious disease susceptible groups based on WiFi matching. It mainly uses user WiFi connection records collected based on mobile devices and anonymous identity code lists of confirmed users. Through record matching, information compression, and construction of a dangerous WiFi database, Calculate the dangerous WiFi coincidence rate of users and other steps to obtain the final list of dangerous users.
  • Figure 1 shows the specific process of the contact tracing pre-screening method for infectious disease susceptible groups based on WiFi matching. The entire embodiment processes the user WiFi connection record data set according to the process shown in Figure 1 and finally obtains the dangerous person screening results.
  • Figure 2 shows the comparison results of the precision rate and recall rate of this method in user groups.
  • the analysis, correction and dimensionality reduction of WiFi connection record data can be done without involving location and other private information, without using
  • the additional equipment also saves computing resources, reduces processing time, improves screening efficiency, and provides another dimension of more comprehensive data support for subsequent searches for high-risk susceptible groups who may have come into contact with confirmed users.

Landscapes

  • Engineering & Computer Science (AREA)
  • Public Health (AREA)
  • Health & Medical Sciences (AREA)
  • Databases & Information Systems (AREA)
  • Medical Informatics (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Epidemiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Security & Cryptography (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法。该方法基于移动设备收集的用户WiFi连接记录与确诊用户匿名身份码名单，经过记录匹配、信息压缩、构造危险WiFi数据库、计算用户危险WiFi重合率等步骤，最终给出用户是否有感染风险的判断。其中所需的数据对于一般的智能移动设备是容易获取的，无需特殊的应用程序。相较于传统的基于GPS与蓝牙的接触追踪方法，该方法在不使用额外设备与敏感数据的基础上提供了另一个维度的信息，运行效率较高，能够帮助更全面高效地开展接触追踪。

Description

一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法 技术领域
本发明涉及传染病易感人群检测，尤其涉及一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法。
背景技术
2020年初爆发的新型冠状病毒疫情对全世界的生产生活造成了极大的冲击，截至2022年3月，全球累计确诊人数已超过4.7亿，累计死亡人数超过600万，世界卫生组织已将2019冠状病毒病(COVID-19)认定为史上最严重的公共卫生时间之一。新冠疫情给全球带来了巨大的生命财产损失，造成了国家、社会、人群间的严重割裂，深刻地改变着世界格局。
在防控疫情的过程中，接触追踪是十分关键的一个步骤。在公共卫生领域，接触者追踪指的是识别可能与受感染者接触的接触者的过程，以及随后收集有关这些接触者的进一步信息的过程。由于包括COVID-19在内的多种传染病都需要通过呼吸道传播、接触传播等方式在人群间进行传播，即发生人与人之间的接触，因此早日发现这些曾与感染者接触过的高危人群对及时隔离潜在的感染者有非常重要的作用。
传统的接触追踪方法主要依赖于对感染者的问卷调查，这种方法高度依赖于受调查者的记忆，可靠度差，同时耗费大量人工，效率低下。随着信息化时代的发展，数字接触追踪技术例如使用智能终端上的应用程序进行接触追踪等逐渐成为解决这一问题的答案。数字接触追踪通常需要借助位置数据，如GPS、WiFi、通信基站、蓝牙信标等，根据使用数据类型又可进一步分为两类方法：
1)使用绝对位置数据的方法：使用绝对的位置信息，如GPS位置、相对于静态的WiFi热点或通信基站的位置等数据，这些数据通常可以表示为地理坐标的形式。这些数据对于传染病接触追踪可能精确率不够高，但是通过大量此类数据能够发现特定用户的行动模式从而进行聚类。
2)使用相对位置数据的方法：使用相对的位置信息，如两台开启了蓝牙的设备的交互数据、同时使用同一台公共交通工具如飞机、火车等的数据，这些数据通常表示为在某一时间点上位于同一区域的数据对。这些数据相对而言有更高的精确率，但一方面需要特定设备(如蓝牙)，另一方面过于关注点与点的互动，而缺乏整体的信息。
两类数据都有了一些实际应用。如中国普遍使用的健康码就利用了用户的GPS序列数据进行接触判断与追踪，而新加坡政府提供的TraceTogether服务就通过强制要求公民下载专门的使用蓝牙信号的应用程序或佩戴专门的蓝牙通信设备来进行接触追踪，但两种方案在应用上都有其缺陷：GPS数据在室内精度有限，且涉及到人的具体位置轨迹等敏感信息；蓝牙点对点接触数据需要大规模使用专业设备，成本高昂不易普及。
针对上述方法的缺点，申请人提出一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法。
发明内容
本发明目的在于对现有应用技术的缺陷加以弥补，通过使用不同类型的数据，提供更全面的传染病易感人群接触追踪方法，该方法可以利用历史传染病确诊患者的WiFi接入记录构造危险WiFi数据库，在获取每日更新的用户WiFi接入记录后给出用户的感染风险，提高寻找易感人群的效率。
本发明的目的通过以下的技术方案实现：
一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法，包括以下步骤：
1)用户匿名身份码记录匹配：通过数据库获取确诊用户唯一匿名身份码数据表，利用hive sql搜索这些id在总数据库中对应的WiFi连接记录，按照日期建立数据表分区，保存为原始数据记录，数据时间跨度总计T天，构造确诊用户WiFi连接数据集R＝{R 1,R 2,…,R T}，其中第i天的记录为R i，总计有k i条记录，则
Figure PCTCN2022088158-appb-000001
其中每条记录r ij＝(gid ij,confirmDay ij,mac ij)，gid ij表示用户唯一匿名身份码、confirmDay ij表示用户确诊日期、mac ij表示用户连接WiFi的mac地址；
2)数据预处理：对步骤(1)中获取的数据集R保存至代码本地并存储，为节省存储空间，建立双向映射字典类，在获取某字符串后，根据内部已存储条目数量针对该字符串建立新的指代符号，将该字符串与指代符号分别作为键与值建立映射；建立用户匿名身份码的代码双向映射字典dict user，记用户总数为N u，将原本的用户统一长度的匿名身份码映射为代号
Figure PCTCN2022088158-appb-000002
其中i u＝1,2,3,…,N u；建立WiFi的代码双向映射字典dict wifi，记WiFi总数为N w，将原本的WiFi的mac地址映射为代号
Figure PCTCN2022088158-appb-000003
其中u w＝1,2,3,…,N w；原始记录数据集映射后得到的记录数据集记为
Figure PCTCN2022088158-appb-000004
3)构造危险WiFi数据库：利用步骤(2)中经过代号映射后的信息数据集
Figure PCTCN2022088158-appb-000005
记危险WiFi 数据库为dict dan，提取
Figure PCTCN2022088158-appb-000006
中确诊用户的WiFi连接记录，取记录中的WiFi代号
Figure PCTCN2022088158-appb-000007
放入dict dan，同时记录该WiFi收入危险WiFi数据库中的日期为当前日期day now；若该WiFi已经存在，则改为更新其收入日期为day now，然后将收入日期
Figure PCTCN2022088158-appb-000008
与当前日期day now之差超过危险WiFi消失阈值thr dan的记录删去；
4)感染风险判断：记每日新获取的用户WiFi连接记录
Figure PCTCN2022088158-appb-000009
将
Figure PCTCN2022088158-appb-000010
按用户进行分组，对每一用户，将其当日连接的所有WiFi与dict dan进行匹配，记重合率r user＝|C user|/|W user|，其中|C user|为该用户当日连接的在dict dan中的WiFi数量，|W user|为该用户当日连接的WiFi总数，若r user高于判断阈值thr user，则判断此用户为危险用户。
进一步地，所述步骤1)用户匿名身份码记录匹配中，使用hive sql工具将确诊用户唯一匿名身份码与全国用户WiFi连接记录总数据库进行匹配，并生成用于后续处理分区确诊用户数据记录表，具体实施方法如下：
1.1)原始数据库包含两部分：1.1.1)确诊用户信息数据表，包括确诊用户匿名身份码gid_confirm、确诊日期confirm_day；1.1.2)全国用户WiFi连接记录数据表，包括用户匿名身份码gid、连接WiFi的mac地址mac、记录生成日期day、记录生成地区area、记录生成时刻hour、记录时刻内WiFi连接次数times，以两个数据库的相同的用户匿名身份码为索引，利用hive sql工具生成确诊用户WiFi连接记录数据表；
其中全国用户WiFi连接记录数据表按地区area与日期day进行分区，在构造目标数据表时，为提高筛选效率、减少筛选时间，需要提前确定目标筛选地区(其地区代码为AREACORE)与目标考察的起止时间(分别为BEGINDAY与ENDDAY)；
1.2)为防止后续处理确诊用户WIFI连接记录数据表因数据量大导致内存溢出，使用hive sql工具将数据表依照日期为索引进行分区处理。
进一步地，所述步骤2)数据预处理中，使用双向映射字典对原始用户匿名身份码与WiFi的mac地址进行压缩与还原，具体实施方法如下：
2.1)构造双向映射字典类TwoWayDict，其特点在于输入后原始未压缩字符串，字典内部形成未压缩字符串到压缩字符串、压缩字符串到未压缩字符串的双向映射，即将未压缩字符串与压缩字符串各作为键与对应的值保存，以在后续处理中节省存储空间与内存，当获取预测结果后，同样根据该双向字典获取原始的未压缩字符串；
2.2)分别将用户匿名身份码与WiFi的mac地址以上述方法建立两个双向映射字典dict user与dict wifi。
进一步地，所述步骤3)构造危险WiFi数据库中，依照日期将历史用户WiFi连接数据中 的mac地址加入危险WiFi数据库，并动态删除过期WiFi以降低假阳性率，具体实施方法如下：
3.1)处理历史数据，提取历史确诊用户WiFi连接记录中的WiFi的mac地址与确诊日期，将其放入危险WiFi数据库dict dan中，dict dan以WiFi的mac地址为键，以确诊日期为值，有重复mac地址的，选择日期最近的保存为值；
3.2)处理每日更新数据，当获取每日更新数据后，将当日确诊患者所连接的WiFi的mac地址以上述方法加入dict dan中，添加完毕后遍历dict dan，如果有mac地址对应的确诊日期与当前日期的差值超过危险WiFi消失阈值thr dan，则将其从危险WiFi数据库中删去。
进一步地，所述步骤4)感染风险判断中，将每日更新的用户WiFi连接数据与危险WiFi数据库进行比对获取重合率，并根据阈值判断是否为危险用户，具体步骤如下：
4.1)获取每日更新的用户WiFi连接记录后，按照用户匿名身份码进行分组，计算其当日连接WiFi与危险WiFi数据库的重合率，用户user的重合率r user计算公式为：
C user＝W user∩W dan
Figure PCTCN2022088158-appb-000011
其中W user为该用户当日连接的WiFi，W dan为危险WiFi数据库，|S|表示集合S的元素个数；
若R user高于危险判断阈值thr user，将其判定为危险用户，应对其进行重点筛查。
本发明一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法，主要通基于移动设备收集的用户WiFi连接记录与确诊用户匿名身份码名单，经过记录匹配、信息压缩、构造危险WiFi数据库、计算用户危险WiFi重合率等步骤得到最终的危险用户名单。本发明对WiFi连接记录数据进行分析、修正、降维，相比于接触追踪方法而言可以在不涉及位置等隐私信息，不使用额外设备的同时节省计算资源，减少处理时间，提高筛查效率，为后续寻找可能与确诊用户发生接触的高危易感人群提供了另一个维度更全面的数据支撑。
附图说明
图1是本发明的基于WiFi匹配的数字接触追踪预筛选的方法流程图；
图2是本发明在实施例用户群中筛选的精确率与召回率比较结果图。
具体实施方式
以下结合附图对本发明的具体实施方法与工作原理作如下详述：
实施例
本实施例对从某地获取的自2020年1月1日至2020年3月20日采集到的用户WiFi连接记录数据以及该时间段内的某呼吸道传染病确诊用户匿名信息。数据集包括的具体变量以及相关数据信息如表1、表2、表3、表4所示：
表1 某地部分用户WiFi连接记录数据
Figure PCTCN2022088158-appb-000012
表2 某地部分用户WiFi连接记录数据字段说明
字段名 数据类型 说明
gid string 用户匿名身份码
hour int 生成记录的时间段
mac string 连接WiFi的mac地址
times int 记录生成时间段内连接该WiFi的次数
day int 记录生成的日期
area int 记录所属地区的区域代码
表3 全国确诊用户数据
confirm_gid confirm_day day
ANDROID-863cc84e5e0548b696eb0c426f249975 20200315 20200315
ANDROID-7af1888075bf4e439a0fefd4b52a865a 20200315 20200315
IOS-5eb8760cc460319162e31f888bfae1ff 20200315 20200315
ANDROID-8d2c876ddf894639a1a80f3ebb695569 20200315 20200315
ANDROID-5c4f0b105585408b9186dba9454ddd58 20200315 20200315
表4 移动设备数据集中wifi_list变量字段解释
字段名 数据类型 说明
confirm_gid string 确诊用户匿名身份码
confirm_day int 该用户确诊日期
day int 该记录所属日期
本实施例中默认传染病易感人群接触追踪预筛选方法的实施数据集即为上述某地用户的WiFi连接数据与某传染病确诊用户数据，方法结果为得到的危险用户列表，其详细实施步骤具体如下：
1)根据用户匿名身份码匹配WiFi连接记录，利用hive sql工具，根据记录日期与所选地区代码从全国用户WiFi连接记录总表中提取子表以提高匹配效率，以用户匿名身份码相同为筛选条件，保留对应记录中的用户匿名身份码、连接WiFi的mac地址、记录生成日期、用户确诊日期字段，并以记录生成日期为分区索引，对生成的确诊用户WiFi连接记录数据表进行分区处理。为了验证预测效果，本实施例中采样了部分健康用户的记录进行同样的处理，健康用户的确诊日期字段做缺省处理。同时在表中添加seed字段，seed等于0表示健康用户，seed等于1表示确诊用户。
2)利用步骤1)中获取的确诊用户WiFi连接记录数据表按照日期依序处理。当读入数据后，分别提取其gid列与mac地址列，分别构造双向映射字典dict user与dict wifi，对gid字符串与mac地址字符串进行压缩，压缩后的数据表如表5所示：
表5 经双向映射字典压缩后数据表
gid mac times quezhen_date seed
u1 w93 13 \N 0
u1 w865054 11 \N 0
u1 w2255432 15 \N 0
u1 w96 9 \N 0
u1 w865074 6 \N 0
3)处理历史数据，考察所有确诊用户连接的WiFi，考虑某传染病存在潜伏期，即感染后表现为无症状直至确诊，判定记录产生日期若在确诊日期T inc内则将该WiFi的mac地址与当天日期加入危险WiFi数据库中，本实施例中取T inc为7天。获得的危险WiFi数据库形式如下：
{'w3706696':datetime.date(2020,3,18),
'w3658960':datetime.date(2020,2,16),
'w3737021':datetime.date(2020,3,11),
'w3834102':datetime.date(2020,3,15),
'w194670':datetime.date(2020,3,20),
'w2787916':datetime.date(2020,2,23),…}
4)判断每日更新数据，将每日更新的用户WiFi连接记录按用户gid进行分组，并汇总每一用户当日扫描的WiFi，计算与危险WiFi的重合率，根据重合率是否高于阈值thr user给出是否为危险用户的判断，本实施例中thr user设为0.1，构造当日的危险用户集合，其形式如下：
{'u46515','u19505','u12248','u35433','u19808',…}
5)根据当日实际确诊的用户信息更新危险WiFi数据库，已经在危险WiFi数据库内的则更新其记录日期。
6)将当前的压缩用户id映射回原始用户匿名身份码，对这些用户进行重点筛查。如步骤4)中用户代码映射结果为：
ANDROID-21010e17edd0437599c2388cd65d130b
ANDROID-c8607f33b36a40b7b17c227867fb290e
ANDROID-7e2f8561d05f468ca553ae9daa55b18c
ANDROID-018570e1daa544f082b655cf80feae54
ANDROID-cb2c78a391ac47229f2aa499d20a2a5f
(7)本实施例中，每日精确率与召回率如图1所示。考虑到传染病筛查相比于精确率，召回率指标更为重要，因此本发明较好地实现了危险人员的筛查。
本发明一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法，主要通基于移动设备收集的用户WiFi连接记录与确诊用户匿名身份码名单，经过记录匹配、信息压缩、构造危险WiFi数据库、计算用户危险WiFi重合率等步骤得到最终的危险用户名单。图1为基于WiFi 匹配的传染病易感人群接触追踪预筛选方法的具体流程。整个实施例按照图1中所示的流程，对用户WiFi连接记录数据集进行处理并最终得出危险人员筛查结果。图2为本方法在用户群中筛选的精确率与召回率比较结果，对WiFi连接记录数据进行分析、修正、降维相比于接触追踪方法而言可以在不涉及位置等隐私信息、不使用额外设备的同时节省计算资源，减少处理时间，提高筛查效率，为后续寻找可能与确诊用户发生接触的高危易感人群提供了另一个维度更全面的数据支撑。
上述实施例只是本发明的举例，尽管为说明目的公开了本发明的最佳实例和附图，但是本领域的技术人员可以理解：在不脱离本发明及所附的权利要求的精神和范围内，各种替换、变化和修改都是可能的。因此，本发明不应局限于最佳实施例和附图所公开的内容。

Claims (5)

  1. 一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法，其特征在于，包括以下步骤：
    1)用户匿名身份码记录匹配：通过数据库获取确诊用户唯一匿名身份码数据表，利用hive sql搜索这些id在总数据库中对应的WiFi连接记录，按照日期建立数据表分区，保存为原始数据记录，数据时间跨度总计T天，构造确诊用户WiFi连接数据集R＝{R 1,R 2,…,R T}，其中第i天的记录为R i，总计有k i条记录，则
    Figure PCTCN2022088158-appb-100001
    其中每条记录r ij＝(gid ij,confirmDay ij,mac ij)，gid ij表示用户唯一匿名身份码、confirmDay ij表示用户确诊日期、mac ij表示用户连接WiFi的mac地址；
    2)数据预处理：对步骤(1)中获取的数据集R保存至代码本地并存储，建立双向映射字典类，在获取任一字符串后，根据内部已存储条目数量针对该字符串建立新的指代符号，将该字符串与指代符号分别作为键与值建立映射；建立用户匿名身份码的代码双向映射字典dict user，记用户总数为N u，将原本的用户统一长度的匿名身份码映射为代号
    Figure PCTCN2022088158-appb-100002
    其中i u＝1,2,3,…,N u；建立WiFi的代码双向映射字典dict wifi，记WiFi总数为N w，将原本的WiFi的mac地址映射为代号
    Figure PCTCN2022088158-appb-100003
    其中u w＝1,2,3,…,N w；原始记录数据集映射后得到的记录数据集记为
    Figure PCTCN2022088158-appb-100004
    3)构造危险WiFi数据库：利用步骤(2)中经过代号映射后的信息数据集
    Figure PCTCN2022088158-appb-100005
    记危险WiFi数据库为dict dan，提取
    Figure PCTCN2022088158-appb-100006
    中确诊用户的WiFi连接记录，取记录中的WiFi代号
    Figure PCTCN2022088158-appb-100007
    放入dict dan，同时记录该WiFi收入危险WiFi数据库中的日期为当前日期day now；若该WiFi已经存在，则改为更新其收入日期为day now，然后将收入日期
    Figure PCTCN2022088158-appb-100008
    与当前日期day now之差超过危险WiFi消失阈值thr dan的记录删去；
    4)感染风险判断：记每日新获取的用户WiFi连接记录
    Figure PCTCN2022088158-appb-100009
    将
    Figure PCTCN2022088158-appb-100010
    按用户进行分组，对每一用户，将其当日连接的所有WiFi与dict dan进行匹配，记重合率r user＝|C user|/|W user|，其中|C user|为该用户当日连接的在dict dan中的WiFi数量，|W user|为该用户当日连接的WiFi总数，若r user高于判断阈值thr user，则判断此用户为危险用户。
  2. 根据权利要求1中所述的一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法，其特征在于所述步骤1)用户匿名身份码记录匹配中，使用hive sql工具将确诊用户唯一匿名 身份码与全国用户WiFi连接记录总数据库进行匹配，并生成用于后续处理分区确诊用户数据记录表，具体实施方法如下：
    1.1)原始数据库包含两部分：1.1.1)确诊用户信息数据表，包括确诊用户匿名身份码gid_confirm、确诊日期confirm_day；1.1.2)全国用户WiFi连接记录数据表，包括用户匿名身份码gid、连接WiFi的mac地址mac、记录生成日期day、记录生成地区area、记录生成时刻hour、记录时刻内WiFi连接次数times，以两个数据库的相同的用户匿名身份码为索引，利用hive sql工具生成确诊用户WiFi连接记录数据表；
    其中全国用户WiFi连接记录数据表按地区area与日期day进行分区，在构造目标数据表时提前确定目标筛选地区与目标考察的起止时间；
    1.2)使用hive sql工具将数据表依照日期为索引进行分区处理。
  3. 根据权利要求1中所述的一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法，其特征在于所述步骤2)数据预处理中，使用双向映射字典对原始用户匿名身份码与WiFi的mac地址进行压缩与还原，具体实施方法如下：
    2.1)构造双向映射字典类TwoWayDict，其特点在于输入后原始未压缩字符串，字典内部形成未压缩字符串到压缩字符串、压缩字符串到未压缩字符串的双向映射，即将未压缩字符串与压缩字符串各作为键与对应的值保存，以在后续处理中节省存储空间与内存，当获取预测结果后，同样根据该双向字典获取原始的未压缩字符串；
    2.2)分别将用户匿名身份码与WiFi的mac地址建立两个双向映射字典dict user与dict wifi。
  4. 根据权利要求1中所述的一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法，其特征在于所述步骤3)构造危险WiFi数据库中，依照日期将历史用户WiFi连接数据中的mac地址加入危险WiFi数据库，并动态删除过期WiFi以降低假阳性率，具体实施方法如下：
    3.1)处理历史数据，提取历史确诊用户WiFi连接记录中的WiFi的mac地址与确诊日期，将其放入危险WiFi数据库dict dan中，dict dan以WiFi的mac地址为键，以确诊日期为值，有重复mac地址的，选择日期最近的保存为值；
    3.2)处理每日更新数据，当获取每日更新数据后，将当日确诊患者所连接的WiFi的mac地址以上述方法加入dict dan中，添加完毕后遍历dict dan，如果有mac地址对应的确诊日期与当前日期的差值超过危险WiFi消失阈值thr dan，则将其从危险WiFi数据库中删去。
  5. 根据权利要求1中所述的一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法，其特征在于所述步骤4)感染风险判断中，将每日更新的用户WiFi连接数据与危险WiFi数据库进行比对获取重合率，并根据阈值判断是否为危险用户，具体步骤如下：
    4.1)获取每日更新的用户WiFi连接记录后，按照用户匿名身份码进行分组，计算其当日连接WiFi与危险WiFi数据库的重合率，用户user的重合率r user计算公式为：
    C user＝W user∩W dan
    Figure PCTCN2022088158-appb-100011
    其中W user为该用户当日连接的WiFi，W dan为危险WiFi数据库，|S|表示集合S的元素个数；
    若R user高于危险判断阈值thr user，将其判定为危险用户，应对其进行重点筛查。
PCT/CN2022/088158 2022-04-14 2022-04-21 一种基于wifi匹配的传染病易感人群接触追踪预筛选方法 Ceased WO2023197350A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
JP2022555088A JP7564565B2 (ja) 2022-04-14 2022-04-21 ＷｉＦｉマッチングに基づく感染症の感受性集団の接触追跡及びプリスクリーニング方法
US17/925,317 US12087451B2 (en) 2022-04-14 2022-04-21 Contact tracing pre-screening method for infectious disease susceptible people based on WiFi matching

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210413620.2 2022-04-14
CN202210413620.2A CN114842980B (zh) 2022-04-14 2022-04-14 一种基于WiFi匹配的传染病易感人群接触追踪预筛选方法

Publications (1)

Publication Number Publication Date
WO2023197350A1 true WO2023197350A1 (zh) 2023-10-19

Family

ID=82565006

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/088158 Ceased WO2023197350A1 (zh) 2022-04-14 2022-04-21 一种基于wifi匹配的传染病易感人群接触追踪预筛选方法

Country Status (4)

Country Link
US (1) US12087451B2 (zh)
JP (1) JP7564565B2 (zh)
CN (1) CN114842980B (zh)
WO (1) WO2023197350A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2025095111A1 (ja) 2023-11-01 2025-05-08 ダイキン工業株式会社 重合体、撥水剤、繊維製品、及び繊維製品の製造方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104778642A (zh) * 2015-03-20 2015-07-15 北京邮电大学 基于WiFi的校园用户数据处理方法、服务器和监测系统
US20190252078A1 (en) * 2018-02-15 2019-08-15 X Development Llc Predicting the spread of contagions
CN111354473A (zh) * 2020-03-20 2020-06-30 天津绿州能源装备有限公司 基于导航地图和手机定位寻找新冠肺炎感染者的应用系统
CN111354472A (zh) * 2020-02-20 2020-06-30 戴建荣 一种传染病传播监测预警系统和方法
CN111446006A (zh) * 2020-04-08 2020-07-24 陈恬慧 一种传染病疫情中密切接触者的追踪方法
WO2021167159A1 (ko) * 2020-02-21 2021-08-26 (주) 티나쓰리디 전염성을 가지는 바이러스의 확진자 정보를 이용하여 바이러스 감염 위험성을 경고하는 방법, 상기 방법을 실행하는 프로그램 및 상기 프로그램이 설치된 장치

Family Cites Families (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10637683B2 (en) * 2016-12-23 2020-04-28 Centurylink Intellectual Property Llc Smart city apparatus, system, and method
JP7422308B2 (ja) 2018-08-08 2024-01-26 パナソニックＩｐマネジメント株式会社 情報提供方法、サーバ、音声認識装置、及び情報提供プログラム
CN110189802B (zh) * 2019-04-28 2023-05-02 万达信息股份有限公司 基于指标存储模型的双向映射队列研究信息系统
US11413425B2 (en) * 2019-05-24 2022-08-16 Neuroenhancement Lab, LLC Device, system, and method for reducing coronasomnia to enhance immunity and immune response
US12308128B2 (en) * 2019-07-23 2025-05-20 The Broad Institute, Inc. Health data aggregation and outbreak modeling
US20210313074A1 (en) * 2020-04-01 2021-10-07 Pricewaterhousecoopers Llp Automatic contact tracing
CN111405470B (zh) * 2020-04-12 2021-12-03 北京牧家科技有限公司 基于移动手机定位算法的聚集人群追踪系统和方法
CN111683341A (zh) * 2020-06-09 2020-09-18 中电达通数据技术股份有限公司 一种基于Wi-Fi数据的疑似感染人员矩阵雷达实现方法
CN111885502B (zh) * 2020-06-28 2021-07-27 华东师范大学 一种保护隐私的疫情防控预警与追溯系统及方法
CN113873441B (zh) * 2020-06-30 2024-02-09 香港科技大学 接触者追踪的方法及相应的处理方法、系统、设备和介质
US20220028562A1 (en) * 2020-07-22 2022-01-27 Pandemic Insights, Inc. Personal and corporate pathogen-risk assessment with pandemic-bio-surveillance multi pathogen systems
US20220037029A1 (en) * 2020-07-29 2022-02-03 T-Mobile Usa, Inc. Infectious disease heat map and prediction model generation
US20220285036A1 (en) * 2020-10-23 2022-09-08 Titan Health & Security Technologies, Inc. Systems and methods for intelligent and privacy-centric contact tracing
CN112786210B (zh) * 2021-01-15 2021-10-08 华南师范大学 一种疫情传播追踪方法及系统
CN113132912B (zh) * 2021-04-12 2022-11-29 东南大学 基于人地数字画像的传染病密接者的双重追溯方法及系统
CN114121302A (zh) * 2021-11-29 2022-03-01 四川恒升信达科技有限公司 一种传染病情防控系统及方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104778642A (zh) * 2015-03-20 2015-07-15 北京邮电大学 基于WiFi的校园用户数据处理方法、服务器和监测系统
US20190252078A1 (en) * 2018-02-15 2019-08-15 X Development Llc Predicting the spread of contagions
CN111354472A (zh) * 2020-02-20 2020-06-30 戴建荣 一种传染病传播监测预警系统和方法
WO2021167159A1 (ko) * 2020-02-21 2021-08-26 (주) 티나쓰리디 전염성을 가지는 바이러스의 확진자 정보를 이용하여 바이러스 감염 위험성을 경고하는 방법, 상기 방법을 실행하는 프로그램 및 상기 프로그램이 설치된 장치
CN111354473A (zh) * 2020-03-20 2020-06-30 天津绿州能源装备有限公司 基于导航地图和手机定位寻找新冠肺炎感染者的应用系统
CN111446006A (zh) * 2020-04-08 2020-07-24 陈恬慧 一种传染病疫情中密切接触者的追踪方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2025095111A1 (ja) 2023-11-01 2025-05-08 ダイキン工業株式会社 重合体、撥水剤、繊維製品、及び繊維製品の製造方法
KR20260074395A (ko) 2023-11-01 2026-05-22 다이킨 고교 가부시키가이샤 중합체, 발수제, 섬유 제품, 및 섬유 제품의 제조 방법

Also Published As

Publication number Publication date
US20240257985A1 (en) 2024-08-01
JP7564565B2 (ja) 2024-10-09
US12087451B2 (en) 2024-09-10
CN114842980B (zh) 2023-07-25
CN114842980A (zh) 2022-08-02
JP2024519625A (ja) 2024-05-21

Similar Documents

Publication Publication Date Title
Moonan et al. Using GIS technology to identify areas of tuberculosis transmission and incidence
CN106960006B (zh) 一种不同轨迹间相似度度量系统及其度量方法
CN113934895A (zh) 一种辅助建立患者主索引的方法
CN103914659A (zh) 基于频率的轨迹抑制数据发布隐私保护的系统及其方法
CN110910991B (zh) 一种医用自动图像处理系统
CN112148843B (zh) 文本处理方法、装置、终端设备和存储介质
CN111326236A (zh) 一种医疗图像自动处理系统
CN113204644B (zh) 一种基于知识图谱的政务百科构建方法
CN113904872A (zh) 一种针对匿名服务网站指纹攻击的特征提取方法及系统
Winkler Approximate string comparator search strategies for very large administrative lists
CN116431625A (zh) 一种地理实体的定位分析方法、装置及计算机设备
CN116881430A (zh) 一种产业链识别方法、装置、电子设备及可读存储介质
Xiao et al. Social ties, spatial migration paradigm, and mental health among two generations of migrants in China
CN117851411B (zh) 患者主索引生成方法及系统
CN114092729A (zh) 基于聚类匿名化与差分隐私保护的异构用电数据发布方法
CN116226103A (zh) 一种基于FPGrowth算法进行政务数据质量检测的方法
CN118656698B (zh) 一种多元历史文化资源的分类方法及设备
WO2017000817A1 (zh) 获取数据之间的匹配关系的方法和装置
CN109885797B (zh) 一种基于多身份空间映射的关系网络构建方法
CN111241164A (zh) 一种中药系统药理学分析平台及分析方法
JP7564565B2 (ja) ＷｉＦｉマッチングに基づく感染症の感受性集団の接触追跡及びプリスクリーニング方法
CN118981509A (zh) 基于多源数据融合的参保扩面人员筛选定位方法及系统
CN110188954A (zh) 基于pop网络的地标可靠性评估方法及装置
CN111581304B (zh) 一种基于社会人口家族关系自动绘制家系图谱的方法
CN120561481A (zh) 一种基于知识图谱模型的群体识别方法、装置和存储介质

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 2022555088

Country of ref document: JP

WWE Wipo information: entry into national phase

Ref document number: 17925317

Country of ref document: US

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

Ref document number: 22936994

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: 22936994

Country of ref document: EP

Kind code of ref document: A1