Method for modeling and predicting spatiotemporal risks of epidemics based on mobile communication data

Abstract

The application discloses a kind of epidemic space-time risk modeling and prediction method based on mobile communication data, comprising: obtaining mobile communication data from communication operator or mobile base station, carry out data mining and analysis, extract user's call time and geographic location information;The space-time contact behavior of individual is analyzed, the contact mode between individuals is mined based on user behavior characteristics, the weight is set according to the activity duration of user in each geographic location, and the interpersonal contact network with space-time attribute is constructed;The dynamic simulation analysis of epidemic risk propagation is carried out to the constructed space-time contact network, and the propagation mechanism is deeply understood;According to the dynamic simulation analysis result, the epidemic risk is evaluated, and the geographic location risk index is given;The risk of user is quantitatively evaluated, and the risk index is proposed, and the risk level is divided, so as to predict the epidemic propagation risk.The application can implement prevention and control measures according to risk level, and improve the effectiveness of high-cost measures.

Description

Technical Field

[0001] This invention relates to the field of epidemic risk prediction technology, and in particular to a method for modeling and predicting the spatiotemporal risk of epidemics based on mobile communication data. Background Technology

[0002] In recent years, the rapid development of big data has driven the advancement of data science. Due to the multi-source heterogeneity, interactivity, and spatiotemporal heterogeneity of the contact network structure in complex social systems, abundant data resources help to improve the network topology, enhance the expressiveness of the network structure, and reveal more features during network modeling. With the rise of data science, data-driven network modeling methods can utilize representative real-world data to make network structures increasingly closer to reality. Therefore, more and more real-world data reflecting spatiotemporal interactions between individuals is being used to enhance the robustness of network models.

[0003] A major challenge in studying data-driven disease transmission processes lies in obtaining the potential contact network structure of the population under study. For diseases with low transmission rates caused by specific transmission methods or chance contact, tracing transmission routes can more accurately describe specific contact patterns between individuals, thus enabling the construction of contact networks. In today's information-driven society, people constantly interact and communicate primarily through smart mobile devices, generating a vast amount of user information records, providing rich data resources for constructing realistic contact networks. This communication data includes the duration of a user's communication at a base station, reflecting the user's geographical location in real time. Frequent population movement and the proximity of interpersonal social behaviors can inadvertently promote the widespread spread of epidemics.

[0004] In studies of epidemic transmission, the spatiotemporal complexity of human contact cannot be fully described by simplified transmission models. Most infectious disease models do not adequately consider how individual behavior changes over time, occasion, and frequency of contact. Current epidemic transmission risk assessments targeting users and base stations directly modeling users monitored from base stations result in excessive spatiotemporal overlap, leading to a high false alarm rate, and fail to provide a quantitative assessment of the epidemic risk level for users and base stations. Summary of the Invention

[0005] The purpose of this invention is to provide a method for modeling and predicting the spatiotemporal risk of epidemics based on mobile communication data, in order to solve the above-mentioned problems.

[0006] The present invention solves the technical problem by adopting the following technical solution:

[0007] A method for modeling and predicting the spatiotemporal risk of epidemics based on mobile communication data includes the following steps:

[0008] Step 1: Obtain mobile communication data from telecommunications operators or mobile base stations, perform data mining and analysis, and extract users' call times and geographical location information;

[0009] Step 2: Analyze the spatiotemporal contact behavior of individuals, mine contact patterns between individuals based on user behavior characteristics, assign weights according to the activity duration of users in each geographical location, and construct an interpersonal contact network with spatiotemporal attributes;

[0010] Step 3: Conduct dynamic simulation analysis of epidemic risk transmission on the constructed spatiotemporal contact network to gain a deeper understanding of the transmission mechanism;

[0011] Step 4: Based on the results of dynamic simulation analysis, assess the risk of the epidemic and provide geographical location risk indicators;

[0012] Step 5: Conduct a quantitative assessment of the user's risk, propose risk indicators, and classify risk levels in order to predict the risk of epidemic transmission.

[0013] Furthermore, the method for extracting the user's call time and geographic location information as described in step 1 includes: communication operators can provide information such as call record details and base station data. When a user makes a call, the base station records the call, and the call record details include the user ID, call start time and duration, and associated base station information. Base station data includes the geographical location information of the base station. Based on this information, a spatiotemporal contact network based on the user, base station, and call duration can be constructed.

[0014] Furthermore, the method for constructing a spatiotemporal interpersonal contact network described in step 2 includes: analyzing user behavioral characteristics, mining user contact patterns at various locations, and assuming spatiotemporal intersection between two users at a given base station if their call durations overlap to a certain extent. Therefore, users can be connected to form a network at each base station based on their call records, where each user ID represents a node in the network. Since users may move to different base stations, an agent-based contact network needs to be constructed across multiple base stations. Additionally, the weight of the edge between two users in the network is measured by the product of their call duration at that base station and their total call duration across all locations, thus constructing a multi-base station interpersonal contact network with spatiotemporal attributes.

[0015] Furthermore, the method for conducting dynamic simulation analysis of epidemic risk transmission described in step 3 is as follows: improve the epidemic transmission model (such as the SIR epidemic model), combine the edge weights with the transmission probability between nodes, simulate the dynamic transmission process of the epidemic, and conduct in-depth research on the dynamic transmission mechanism of epidemic risk.

[0016] Furthermore, the risk assessment of the epidemic described in step 4, which provides a geographical location risk index, is conducted as follows: based on the dynamic simulation results, the time it takes for the epidemic to reach the base station and the number of infected individuals in the base station are calculated, and the risk level of the geographical location of the base station is assessed.

[0017] Furthermore, step 5 involves quantitatively assessing the user's risk, proposing risk indicators, and classifying risk levels to predict the risk of epidemic transmission. The method is as follows: simulate the dynamic transmission process of the epidemic, calculate the risk of each person being infected with and spreading the epidemic from the numerical simulation results, and classify the user's risk level to predict the risk of epidemic transmission.

[0018] Beneficial effects:

[0019] This invention utilizes complex network theory and data mining techniques to model and predict the spatiotemporal transmission risk of epidemics based on mobile communication data. Existing technologies predict epidemic risk based on users monitored by base stations, but the spatiotemporal overlap of users is often too large, resulting in a high false alarm rate and making it difficult to quantitatively assess the risk levels of base stations and users. In contrast, this invention mines user communication data and location information combined with the geographical location of base stations to deeply study users' spatiotemporal interaction behavior and fully utilizes call duration and location data to construct a weighted contact network of multiple base stations. Therefore, it can predict and assess the spatiotemporal transmission risk in the early stages of an epidemic, not only more accurately locating users and base stations with higher risk levels but also helping those developing prevention and control measures to implement corresponding measures based on the estimated risk, thereby significantly improving the effectiveness of prevention and control measures. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the method flow described in this invention;

[0021] Figure 2 This is an example diagram of an interpersonal contact network with spatiotemporal attributes constructed based on the abstract representation of users and base stations using mobile communication data in an embodiment of the present invention.

[0022] Figure 3 This is a visualization of base station risk prediction based on the arrival time of an epidemic, as described in an embodiment of the present invention.

[0023] Figure 4 This is a visualization of base station risk prediction based on the total number of infections generated by an epidemic, as described in an embodiment of the present invention.

[0024] Figure 5 This is a visualization of the risk of epidemic transmission among users in an embodiment of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] This invention provides a method for modeling and predicting the spatiotemporal risk of epidemics based on mobile communication data. It abstracts the contacts formed by the spatiotemporal intersection between users using nodes and edges, and then simulates the dynamic propagation of epidemics on the established multi-base station complex network model. This method can predict the geographical location risk of base stations and the risk to individual users in the early stages of propagation, thereby improving the effectiveness of prevention and control measures.

[0027] like Figure 1 As shown, the present invention provides a method for modeling and predicting the spatiotemporal risk of an epidemic based on mobile communication data, comprising the following steps:

[0028] Step 1: Acquire mobile communication data, perform data mining and analysis, and determine the nodes and connections in the contact network.

[0029] Based on mobile communication data obtained from telecommunications operators (such as data from a week or a day), since the data contains users' communication records and associated base stations, as well as the geographical location information of the base stations, it is first necessary to determine the number of users and call duration in each base station. After the intersection of call durations between users reaches a certain proportion, it is assumed that there is a spatiotemporal intersection between users, that is, a connection is generated.

[0030] Step 2: Analyze users' communication data, uncover contact patterns between individuals, and construct a weighted interpersonal contact network with spatiotemporal attributes.

[0031] In-depth analysis of mobile communication data is conducted to uncover users' spatiotemporal behavior models, namely the proportion of a user's call duration at each base station to the total call duration during the research period. Therefore, the weight of the connection between users in the same base station is measured by the product of the proportions of their call durations. Figure 2 This is an example diagram of an interpersonal contact network with spatiotemporal attributes, constructed by abstracting users and base stations based on mobile communication data.

[0032] Step 3: Based on a complex network model with multiple base stations, simulate the dynamic spread of an epidemic and explore its transmission mechanism.

[0033] Based on the constructed multi-base station interpersonal contact network, the risk of epidemic transmission can be modeled. The main application is epidemic transmission simulation. By combining edge weights, the transmission probability part in the model is improved, thereby simulating the spatiotemporal transmission process of epidemics and exploring the dynamic transmission mechanism of epidemics.

[0034] Step 4: Based on the simulation results, assess the risks of the base station and provide risk indicators.

[0035] Based on dynamic simulation results, the spatiotemporal transmission risk of an epidemic can be assessed. First, the risk at different geographical locations is quantitatively assessed, primarily by calculating the time the epidemic arrives at each base station and the total number of infections after the transmission process ends. These are used as risk indicators for each location, and then the risk levels of each base station are classified. Figure 3 and Figure 4 This is a visual representation of the risk assessment of base stations.

[0036] Step 5: Based on the analysis results, quantitatively assess the user's risk, propose risk indicators, and classify risk levels.

[0037] User behavior analysis is performed on the dynamic simulation results to quantitatively assess users' infection risk and the risk of spreading an epidemic. This involves calculating the infection probability of individuals within the network and the effective reproduction number during transmission as risk indicators for users. These values ​​are then normalized to classify users into risk levels. Figure 5 A visual representation of the risk of epidemic transmission to users.

[0038] The spatiotemporal heterogeneity and high complexity of epidemic transmission pose significant challenges to the study of spatiotemporal transmission risks. Many epidemics spread precisely because of human movement and contact; studying the spatiotemporal contact patterns of groups helps to deepen our understanding of epidemic transmission mechanisms. In the era of big data, how to utilize the massive amounts of data generated by wireless communication and mobile internet technologies to model and predict the spatiotemporal transmission risks of epidemics is of great significance for social management and public health epidemic prevention and control. Therefore, this invention will utilize user communication time and geographical location recorded in mobile communication data to obtain people's movement and contact patterns with individual-based spatiotemporal granularity, construct a contact network, predict the risk of epidemic transmission, and then propose effective prevention and control measures.

[0039] Through the above steps, a method for modeling and predicting the spatiotemporal risk of epidemics based on mobile communication data can be constructed. This method is used by policymakers to predict the risk of epidemic transmission caused by population movement before designing and implementing prevention and control measures. Given the rapidly increasing scale and growing spatiotemporal heterogeneity of epidemics, existing methods for predicting transmission risks face the problem of being unable to quantitatively assess geographical location risk and individual risk. The method of this invention models and predicts the spatiotemporal transmission risk of epidemics from the perspective of the cross-integration of complex network theory and data mining. In this technology, we utilize complex network theory and data mining techniques to model and quantitatively assess the spatiotemporal transmission risk of epidemics. This method allows for a preliminary assessment of the spatiotemporal transmission risk of epidemics after an outbreak, before the formulation and implementation of prevention and control measures. This enables policymakers to predict the potential risks of epidemic transmission based on geographical location and among users before the implementation of prevention and control measures, thereby allowing for targeted implementation of prevention and control policies and significantly improving the effectiveness of high-cost measures.

[0040] This invention is applicable to policymakers who can predict the potential risks of epidemic transmission between geographical locations and users before implementing prevention and control measures, thereby enabling them to implement prevention and control measures in a targeted manner according to risk levels and improve the effectiveness of high-cost measures.

[0041] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for modeling and predicting the spatiotemporal risk of epidemics based on mobile communication data, characterized in that, Includes the following steps: Step 1: Obtain mobile communication data from telecommunications operators or mobile base stations, perform data mining and analysis, and extract users' call times and geographical location information; Step 2: Analyze the spatiotemporal contact behavior of individuals, mine contact patterns between individuals based on user behavior characteristics, assign weights according to the activity duration of users in each geographical location, and construct an interpersonal contact network with spatiotemporal attributes; Step 3: Conduct dynamic simulation analysis of epidemic risk transmission on the constructed spatiotemporal contact network to gain a deeper understanding of the transmission mechanism; Step 4: Based on the results of dynamic simulation analysis, assess the risk of the epidemic and provide geographical location risk indicators; Step 5: Conduct a quantitative assessment of the user's risk, propose risk indicators, and classify risk levels in order to predict the risk of epidemic transmission; The method for constructing a spatiotemporal interpersonal contact network as described in step 2 includes: analyzing user behavior characteristics, mining user contact patterns in various locations, and building a network by connecting users in each base station based on their call start time and call duration, where each user ID represents a node in the network; users will move to different base stations, requiring the construction of an agent-based contact network across multiple base stations; the weight of the edge between two users in the network is measured by the product of their call duration at that base station and their call duration across all locations, thereby constructing a multi-base station interpersonal contact network with spatiotemporal attributes; The method for conducting dynamic simulation analysis of epidemic risk transmission described in step 3 is as follows: improve the epidemic transmission model by combining the weight of the edges with the transmission probability between nodes to simulate the dynamic transmission process of the epidemic and conduct in-depth research on the dynamic transmission mechanism of epidemic risk.

2. The method for modeling and predicting the spatiotemporal risk of an epidemic based on mobile communication data according to claim 1, characterized in that, The method for extracting the user's call time and geographic location information as described in step 1 includes: when a user makes a call, the base station records the user's call, and the communication record details include the user ID, call start time and duration, and associated base station information; this information is used to construct a spatiotemporal contact network based on the user, the base station, and the call duration.

3. The method for modeling and predicting the spatiotemporal risk of an epidemic based on mobile communication data according to claim 1, characterized in that, The risk assessment of the epidemic described in step 4, which provides a geographical location risk index, is conducted as follows: Based on the dynamic simulation results, the time when the epidemic arrives at the base station and the number of infected persons in the base station are calculated, and the risk level of the geographical location of the base station is assessed.

4. The method for modeling and predicting the spatiotemporal risk of an epidemic based on mobile communication data according to claim 1, characterized in that, The quantitative assessment of user risk described in step 5, proposing risk indicators, and classifying risk levels in order to predict the risk of epidemic transmission, is carried out by the following method: simulating the dynamic transmission process of an epidemic, calculating the risk of each person being infected with and spreading the epidemic from the numerical simulation results, and classifying the user's risk level, thereby predicting the risk of epidemic transmission.

Images