CN113792763B - Social group behavior identification method, computer device and storage medium based on electromagnetic spectrum data mining - Google Patents

Abstract

The invention discloses a social group behavior identification method based on electromagnetic spectrum data mining, a computer device and a storage medium, wherein the social group behavior identification method comprises the following steps: the method comprises the steps of obtaining the frequency band occupancy rate and the characteristic data of social group behaviors, extracting an optimal characteristic subset from the characteristic data, determining the correlation between the frequency band occupancy rate and the optimal characteristic subset, carrying out time-varying behavior analysis on the optimal characteristic subset to obtain a first time sequence, and constructing a social group behavior prediction model for social group behavior recognition according to the correlation and the first time sequence. The invention can complete the identification only by analyzing the physical layer, does not interfere the communication process and the radio environment in the monitoring process, does not need to pass through a communication operator or a user on the premise of legal execution, does not have the risk of invading the privacy and has small implementation difficulty. The method is widely applied to the technical field of data mining.

Description

Social group behavior identification method, computer device and storage medium based on electromagnetic spectrum data mining

technical field

The invention relates to the technical field of data mining, in particular to a social group behavior identification method, computer device and storage medium based on electromagnetic spectrum data mining.

Background technique

Social group behaviors refer to the behaviors displayed by the social activities of the crowd, such as Spring Festival travel, watching sports, disaster relief, subway evacuation and other behaviors can be classified as social group behaviors. Identifying the characteristics of social group behaviors is helpful. Carry out social management and decision-making work such as people flow management and resource allocation. In modern society, people generally use mobile phones, tablet computers, wearable devices and other devices that use electromagnetic wireless communication. Because these devices are carried around and emit signals almost all the time, they are used in the identification of social group behaviors, such as , to link and analyze the data generated by mobile communication with the distribution of social group behavior. However, the prior art needs to obtain the data generated by the mobile phone communication from the operator, and may need to parse the data generated by the mobile phone communication to confirm the content, which involves issues such as data ownership and privacy, which makes the prior art face greater challenges. limitations.

SUMMARY OF THE INVENTION

In view of at least one of the above technical problems, the purpose of the present invention is to provide a social group behavior identification method, computer device and storage medium based on electromagnetic spectrum data mining.

On the one hand, an embodiment of the present invention includes a method for identifying social group behavior based on electromagnetic spectrum data mining, including:

obtaining frequency band occupancy for radio communications conducted in the first area;

obtaining characteristic data of social group behavior in the first area;

extracting the optimal feature subset from the feature data;

Determine the correlation between the frequency band occupancy and the optimal feature subset by using a geographically weighted regression model;

Perform time-varying behavior analysis on the optimal feature subset by using a time series model to obtain a first time series;

Building a social group behavior prediction model by using a deep learning model according to the correlation between the frequency band occupancy and the optimal feature subset and the first time series;

The social group behavior identification is performed using the social group behavior prediction model.

Further, the acquiring the frequency band occupancy of the radio communication performed in the first area includes:

Determine the position of the first area where the signal source is located by using the AOA/TDOA algorithm;

In the first area, electromagnetic spectrum monitoring data is obtained by monitoring the mobile wireless communication device;

The frequency band occupancy is calculated according to the electromagnetic spectrum monitoring data.

Further, the obtaining characteristic data of social group behavior in the first area includes:

measuring crowd movement track data in the first area;

Correlation analysis, factor analysis or cluster analysis is performed on the crowd activity track data, and the obtained result is used as the characteristic data.

Further, extracting the optimal feature subset from the feature data includes:

Dimensionality reduction processing is performed on the feature data using principal component analysis to obtain the optimal feature subset.

Further, according to the correlation between the frequency band occupancy and the optimal feature subset and the first time series, using a deep learning model to build a social group behavior prediction model, including:

determining a second time series corresponding to the first time series according to the correlation between the frequency band occupancy and the optimal feature subset;

Using the second time series as the input of the deep learning model, and using the first time series as the expected output of the deep learning model, train the deep learning model;

The trained deep learning model is used as the social group behavior prediction model.

Further, the deep learning model is a recurrent neural network, a graph neural network, a decision tree model, a random forest model, a Markov model or an ARMA model.

Further, the social group behavior identification method based on electromagnetic spectrum data mining also includes:

Through a geographic information system, the correlation between the frequency band occupancy and the optimal feature subset is visualized.

Further, the use of the social group behavior prediction model to identify social group behaviors includes:

Obtain frequency band occupancy for radio communications in a specific area over a period of time;

inputting the frequency band occupancy into the social group behavior prediction model;

According to the output result of the social group behavior prediction model, the characteristics of the social group behavior are determined.

On the other hand, an embodiment of the present invention further includes a computer device, including a memory and a processor, where the memory is configured to store at least one program, and the processor is configured to load the at least one program to execute an electromagnetic spectrum data mining-based algorithm. Methods for identifying social group behavior.

On the other hand, an embodiment of the present invention further includes a storage medium, in which a processor-executable program is stored, and when the processor-executable program is executed by the processor, the program is used to execute a social property based on electromagnetic spectrum data mining. Group behavior identification methods.

The beneficial effects of the present invention are: the social group behavior identification method based on electromagnetic spectrum data mining in the embodiment can identify the characteristics of social group behavior by analyzing the frequency band occupancy degree of radio communication used by the crowd, and these information It can be used for social management and decision-making work such as people flow management and resource allocation to improve the efficiency of these work. Compared with the prior art, the social group behavior identification method based on electromagnetic spectrum data mining in the embodiment does not need to rely on identifying the data content contained in the wireless signal, and only needs to carry out the analysis at the physical level to complete the identification, and the monitoring process is not required. It will cause interference to the communication process and the radio environment. Under the premise of legality, it does not need to go through communication operators or users, and there is no risk of privacy violation due to interception of data, so the implementation difficulty is small.

Description of drawings

1 is a flowchart of a social group behavior identification method based on electromagnetic spectrum data mining in an embodiment;

2 is a schematic diagram of a social group behavior identification method based on electromagnetic spectrum data mining in an embodiment;

3 is a schematic diagram of a social group behavior distribution monitoring platform based on electromagnetic spectrum data mining in an embodiment;

4 is a schematic diagram of a main working frequency band of a first mobile communication operator in an embodiment;

5 is a schematic diagram of a main working frequency band of a second mobile communication operator in an embodiment;

6 is a schematic diagram of a main working frequency band of a third mobile communication operator in an embodiment;

FIG. 7 is a schematic diagram of the effect of applying the method for identifying social group behavior based on electromagnetic spectrum data mining to several regions in the embodiment.

Detailed ways

In this embodiment, referring to FIG. 1 , the method for identifying social group behavior based on electromagnetic spectrum data mining includes the following steps:

S1. Obtain the frequency band occupancy of the radio communication performed in the first area;

S2. Obtain characteristic data of social group behavior in the first area;

S3. Extract the optimal feature subset from the feature data;

S4. Determine the correlation between the frequency band occupancy and the optimal feature subset through a geographically weighted regression model;

S5. Perform time-varying behavior analysis on the optimal feature subset through the time series model to obtain the first time series;

S6. According to the correlation between the frequency band occupancy and the optimal feature subset and the first time series, use the deep learning model to construct a social group behavior prediction model;

S7. Use a social group behavior prediction model to identify social group behavior.

The principle of steps S1-S7 is shown in FIG. 2 .

Step S1, that is, the step of obtaining the frequency band occupancy of the radio communication performed in the first area, specifically includes the following steps:

S101. Determine the position of the first area where the signal source is located through the AOA/TDOA algorithm;

S102. In the first area, obtain electromagnetic spectrum monitoring data by monitoring the mobile wireless communication device;

S103. Calculate frequency band occupancy according to electromagnetic spectrum monitoring data.

Before performing steps S101-S103, data preprocessing can be performed first, the existing spectrum data can be converted into a format that is easy to calculate, necessary information can be retained, and redundant information can be eliminated. Subsequent calculation speed. The process and principle of steps S101-S103 are: in step S101, the obtained spectrum data is classified into regions, the spatial position of the signal source is located, and the parameters related to the signal source are adjusted through the AOA/TDOA algorithm, and combined with the assistance of geographic information, it can be obtained. Spectrum data belonging to a specific area, so as to obtain high-precision regional social group behavior. The signal source can be mobile phones, tablet computers, Bluetooth headsets, smart bracelets and other devices that use electromagnetic wireless communication. The first area refers to A specific area, such as the area where a specific group of people are using mobile phones, tablet computers, Bluetooth headsets, and smart bracelets; in step S102, the electromagnetic spectrum monitoring data may be obtained through a communication operator, or it may not be operated through communication It uses the instruments set up by a third party to monitor mobile wireless communication equipment to obtain electromagnetic spectrum monitoring data, and calculates the frequency band occupancy in a specific area, that is, the spectrum data in the first area.

Step S2, that is, the step of acquiring characteristic data of social group behavior in the first area, specifically includes the following steps:

S201. Measure crowd activity track data in the first area;

S202. Perform correlation analysis, factor analysis or cluster analysis on the crowd activity track data, and use the obtained results as feature data.

The principle of steps S201-S202 is that: the satellite positioning system or base station positioning system installed on the mobile phone can be used to pre-measure the data of the social group behavior in a specific area, such as crowd activity trajectory data, and then perform correlation analysis, factor analysis, and cluster analysis. By extracting characteristic data significantly related to social group behavior, we can obtain spectral data features significantly related to social group behavior.

Step S3, that is, the step of extracting the optimal feature subset from the feature data, specifically includes the following steps:

S301. Use principal component analysis to perform dimensionality reduction processing on the feature data to obtain an optimal feature subset.

The optimal feature subset is obtained by further reducing the dimension of the feature data by using the principal component analysis method. In the subsequent steps, the optimal feature subset is used instead of the feature data, which can reduce the calculation amount of the training model.

In step S4, a correlation analysis is performed on the frequency band occupancy and the optimal feature subset through a geographically weighted regression model, so as to determine the correlation between the frequency band occupancy and the optimal feature subset. Since the optimal feature subset is extracted from the feature data, the optimal feature subset contains the feature information of social group behavior, and the frequency band occupancy is measured from the radio communication used by the population, including The phenomenon information of social group behavior, so the correlation between the frequency band occupancy and the optimal feature subset represents the corresponding relationship between the phenomenon information and characteristic information of social group behavior. In step S4, through the geographic information system, the correlation between the frequency band occupancy and the optimal feature subset can also be visualized.

In step S5, a time-varying behavior analysis is performed on the optimal feature subset through a time-series model to obtain a first time-series. In step S5, a time series model is used to model the behavior of social groups in a specific area, and time-varying behavior analysis is performed. The selection of a specific area can use the city POI data, that is, select a geographical location with social significance, such as shopping malls, schools, attractions, etc. The obtained first time series includes a plurality of data with the same properties as the optimal feature subset.

Step S6 is the step of constructing a social group behavior prediction model by using a deep learning model according to the correlation between the frequency band occupancy and the optimal feature subset and the first time series, which specifically includes the following steps:

S601. Determine a second time series corresponding to the first time series according to the correlation between the frequency band occupancy and the optimal feature subset;

S602. Use the second time series as the input of the deep learning model, and use the first time series as the expected output of the deep learning model, and train the deep learning model;

S603. Use the trained deep learning model as a social group behavior prediction model.

In step S601, according to the data included in the first time series with the same properties as the optimal feature subset, and the correlation between the frequency band occupancy and the optimal feature subset, a plurality of properties and frequency band occupancy can be obtained. The same data that make up the second time series. According to the properties of the data contained in the first time series and the second time series, the second time series can be used as the input of the deep learning model, and the first time series can be used as the expected output of the deep learning model to train the deep learning model. Specifically, a model such as a recurrent neural network (RNN), a graph neural network (GNN), a decision tree model, a random forest model, a Markov model or an ARMA model can be used as the deep learning model. When training the deep learning model, the data in the second time series can be input to the deep learning model, processed by the deep learning model and output, and the output result of the deep learning model is compared with the expected output, and the depth is adjusted according to the error. The parameters of the learning model are transferred in reverse, and the training of the deep learning model is stopped until the error between the output result of the deep learning model and the expected output is less than the preset threshold.

The trained deep learning model is the social group behavior prediction model in this embodiment, and the obtained social group behavior prediction model has the ability to identify the behavior characteristics of the social group according to the frequency band occupancy.

Step S7, that is, using the social group behavior prediction model to identify the social group behavior, specifically includes the following steps:

S701. Obtain the frequency band occupancy of radio communications in a specific area for a period of time;

S702. Input the frequency band occupancy into the social group behavior prediction model;

S703. Determine the characteristics of social group behavior according to the output result of the social group behavior prediction model.

The specific area in step S701 may refer to a city, a country or even the whole world, and a period of time may refer to a specific time such as a specific year, month, day, or the duration of an event. For example, when step S701 is performed, the time and place of the occurrence of short-term major events can be selected for analysis, such as a concert in a certain city, traffic congestion on a certain road, etc., to observe the changes of electromagnetic data and social group behavior in these scenarios; or Select a long special period, such as the global infectious disease pandemic period, the urban heat wave duration period, etc. for analysis.

By executing steps S701-S703, the characteristics of the social group behavior of the group can be obtained according to the frequency band occupancy generated when the group uses a wireless communication device such as a mobile phone in a region for a period of time, wherein the content of the characteristics of the social group behavior is the same as that of the group. When performing steps S1-S6 training to obtain a social group behavior prediction model, the characteristic data used in step S2 is the same, which can be the population size growth rate, the crowd movement trajectory trend, etc., based on the identification results of the characteristics of the social group behavior , put forward appropriate opinions and suggestions, such as making arrangements for people flow diversion and decision-making. Analyzing social group behavior through electromagnetic spectrum data can be used for a variety of purposes. If the accuracy of the analysis is accurate enough, it can be used to search and rescue people trapped in disaster areas after earthquakes and other disasters occur. By accumulating long-term electromagnetic spectrum data in a certain area, when a major emergency occurs, the impact of the event can be analyzed. Similarly, if it is detected that the electromagnetic spectrum data of a certain area has a large-scale abnormality, the area can be investigated to determine whether a special event has occurred.

The social group behavior identification method in this embodiment can be applied to the monitoring platform shown in FIG. 3 . The social group behavior distribution monitoring platform based on electromagnetic spectrum data mining combines time domain, frequency domain and spatial domain multi-dimensional information to analyze the spectrum data. An analysis is performed that provides prior knowledge about the radio. Through the construction and distribution of this platform, radio regulatory authorities can achieve effective spectrum management, and by calculating the occupancy of the target frequency band, the distribution of social group behaviors in the region can be reflected.

Referring to Fig. 4, Fig. 5, Fig. 6, in the present embodiment, select the main working frequency bands of these three mobile communication operators of the first mobile communication operator in China, the second mobile communication operator and the third mobile communication operator as target frequency bands to illustrate the application of social group behavior identification methods. Among them, Figure 4 shows the main working frequency band of the first mobile communication operator, Figure 5 shows the main working frequency band of the second mobile communication operator, and Figure 6 shows the main working frequency band of the third mobile communication operator.

The social group behavior identification method in this embodiment is applied to Wenchang City, Sanya City, Lingshui Autonomous County, Haikou City, Lingao County, Dongfang City and other regions in Hainan Province, China. Identify group behavior. Step S1 of the social group behavior identification method in this embodiment is performed by monitoring the frequency band occupancy of the nearby target frequency band at the monitoring point in the above-mentioned area. Specifically, the monitoring points located in Haikou City, Lingao County, Dongfang City and other areas are mainly located in residential quarters, dormitories and other residences, and the monitoring points located in Wenchang City, Sanya City and Lingshui Autonomous County are mainly located in such as Wenchang Longlou, Sanya Jiefang Road, Lingshui Autonomous County Center, etc. Different color depths are used to represent the occupancy of the target frequency band in the corresponding region. The darker the color, the higher the utilization of the target frequency band in the region, which further reflects the greater social group behavior in the region.

The social group behavior identification method in this embodiment is applied to the above regions, and the obtained analysis result is shown in FIG. 7 . As can be seen from Figure 7, the frequency band occupancy of Wenchang City, Sanya City, and Lingshui Autonomous County is slightly higher than that of Haikou City, Lingao County, and Dongfang City. According to the principle of the social group behavior identification method in this embodiment, it can be considered that the social group characteristics of Wenchang City, Sanya City, and Lingshui Autonomous County shown in FIG. 7 are higher than those of Haikou City, Lingao County and Dongfang City. Features are obvious. Since the monitoring points in Haikou City, Lingao County, Dongfang City and other areas are mainly located in residential quarters, dormitories and other residences, the social group behavior in these places is relatively inactive, while the monitoring points in Wenchang City, Sanya City, and Lingshui Autonomous County are mainly located in In places such as Wenchang Longlou, Sanya Jiefang Road, Lingshui Autonomous County Center, etc., the social group behaviors in these places are relatively active, so the conclusion drawn according to the social group behavior identification method in this embodiment is consistent with the actual situation. , which confirms the feasibility of the social group behavior identification method in this embodiment.

A computer program can be written according to this method for identifying social group behavior based on electromagnetic spectrum data mining, and the computer program can be written into the memory of the computer device or in an independent storage medium, and after the computer program is read out, the processor can be instructed to execute the method based on electromagnetic spectrum. A social group behavior identification method for spectrum data mining, so as to achieve the same technical effect as the method embodiment.

It should be noted that, unless otherwise specified, when a feature is called "fixed" or "connected" to another feature, it can be directly fixed or connected to another feature, or it can be indirectly fixed or connected to another feature. on a feature. In addition, descriptions such as upper, lower, left, right, etc. used in the present disclosure are only relative to the mutual positional relationship of each component of the present disclosure in the accompanying drawings. As used in this disclosure, the singular forms "a," "the," and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise. Also, unless otherwise defined, all technical and scientific terms used in this embodiment have the same meaning as commonly understood by those skilled in the art. The terms used in the description of the embodiments are only used to describe specific embodiments, rather than to limit the present invention. As used in this example, the term "and/or" includes any combination of one or more of the associated listed items.

It will be understood that, although the terms first, second, third, etc. may be used in this disclosure to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish elements of the same type from one another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the present disclosure. The use of any and all examples or exemplary language ("for example," "such as," etc.) provided in this embodiment is intended only to better illustrate embodiments of the invention and does not detract from the scope of the invention unless otherwise requested impose restrictions.

It should be appreciated that embodiments of the present invention may be implemented or implemented by computer hardware, a combination of hardware and software, or by computer instructions stored in non-transitory computer readable memory. The method can be implemented in a computer program using standard programming techniques - including a non-transitory computer-readable storage medium configured with a computer program, wherein the storage medium so configured causes the computer to operate in a specific and predefined manner - according to the specific Methods and figures described in the Examples. Each program may be implemented in a high-level procedural or object-oriented programming language to communicate with a computer system. However, if desired, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. Furthermore, the program can be run on a programmed application specific integrated circuit for this purpose.

Furthermore, the operations of the processes described in this embodiment may be performed in any suitable order unless otherwise indicated by this embodiment or otherwise clearly contradicted by context. The processes described in this embodiment (or variations and/or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be executed as code collectively executing on one or more processors (eg, executable instructions, one or more computer programs, or one or more applications), implemented in hardware, or a combination thereof. The computer program includes a plurality of instructions executable by one or more processors.

Further, the methods may be implemented in any type of computing platform operably connected to a suitable, including but not limited to personal computer, minicomputer, mainframe, workstation, network or distributed computing environment, stand-alone or integrated computer platform, or communicate with charged particle tools or other imaging devices, etc. Aspects of the invention may be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, an optically read and/or written storage medium, RAM, ROM, etc., such that it can be read by a programmable computer, when a storage medium or device is read by a computer, it can be used to configure and operate the computer to perform the processes described herein. Furthermore, the machine-readable code, or portions thereof, may be transmitted over wired or wireless networks. The invention described in this embodiment includes these and other various types of non-transitory computer-readable storage media when such media includes instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor. The invention also includes the computer itself when programmed according to the methods and techniques described herein.

A computer program can be applied to input data to perform the functions described in this embodiment to transform the input data to generate output data for storage to non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the invention, the transformed data represents physical and tangible objects, including specific visual depictions of physical and tangible objects produced on the display.

The above are only preferred embodiments of the present invention, and the present invention is not limited to the above-mentioned embodiments, as long as it achieves the technical effect of the present invention by the same means, all within the spirit and principle of the present invention, do Any modification, equivalent replacement, improvement, etc., should be included within the protection scope of the present invention. Various modifications and changes can be made to its technical solutions and/or implementations within the protection scope of the present invention.

Claims (9)

1. A social group behavior identification method based on electromagnetic spectrum data mining is characterized by comprising the following steps:

acquiring a frequency band occupancy rate of radio communication performed in a first area;

acquiring characteristic data of social group behaviors in the first area;

extracting an optimal feature subset from the feature data;

determining a correlation between the frequency band occupancy and the optimal feature subset through a geographical weighted regression model;

performing time-varying behavior analysis on the optimal feature subset through a time series model to obtain a first time series;

according to the correlation between the frequency band occupancy and the optimal feature subset and the first time sequence, a deep learning model is used for constructing a social group behavior prediction model;

performing social group behavior recognition by using the social group behavior prediction model;

the acquiring characteristic data of the social group behaviors in the first area comprises the following steps:

measuring crowd activity trajectory data in the first region;

and carrying out correlation analysis, factor analysis or clustering analysis on the crowd activity track data, and taking the obtained result as the characteristic data.

2. The method for social group behavior recognition based on electromagnetic spectrum data mining as claimed in claim 1, wherein the obtaining of the frequency band occupancy of the radio communication performed in the first area comprises:

determining the position of the first area where the signal source is located through an AOA/TDOA algorithm;

acquiring electromagnetic spectrum monitoring data by monitoring a mobile wireless communication device in the first area;

and calculating the frequency band occupancy rate according to the electromagnetic spectrum monitoring data.

3. The method for identifying social group behaviors based on electromagnetic spectrum data mining according to claim 1, wherein the extracting an optimal feature subset from the feature data includes:

and performing dimensionality reduction on the feature data by using a principal component analysis method to obtain the optimal feature subset.

4. The method for identifying social group behaviors based on electromagnetic spectrum data mining as claimed in claim 1, wherein the constructing a social group behavior prediction model using a deep learning model according to the correlation between the frequency band occupancy and the optimal feature subset and the first time series comprises:

determining a second time sequence corresponding to the first time sequence according to the correlation between the frequency band occupancy and the optimal feature subset;

training the deep learning model by taking the second time series as the input of the deep learning model and taking the first time series as the expected output of the deep learning model;

and taking the trained deep learning model as the social group behavior prediction model.

5. A social group behavior recognition method based on electromagnetic spectrum data mining according to claim 4, characterized in that the deep learning model is a recurrent neural network, a graph neural network, a decision tree model, a random forest model, a Markov model or an ARMA model.

6. The method for social group behavior recognition based on electromagnetic spectrum data mining as claimed in claim 4 or 5, further comprising:

and visualizing the correlation between the occupancy rate of the frequency band and the optimal feature subset through a geographic information system.

7. The method for social group behavior recognition based on electromagnetic spectrum data mining as claimed in claim 1, wherein the social group behavior recognition using the social group behavior prediction model comprises:

acquiring the frequency band occupancy rate of radio communication performed in a specific area within a period of time;

inputting the frequency band occupancy into the social group behavior prediction model;

and determining the characteristics of the social group behaviors according to the output result of the social group behavior prediction model.

8. A computer apparatus comprising a memory for storing at least one program and a processor for loading the at least one program to perform the method for electromagnetic spectrum data mining based social group behavior recognition of any of claims 1-7.

9. A storage medium having stored therein a processor-executable program, wherein the processor-executable program, when executed by a processor, is configured to perform the method for identifying social group behaviors based on electromagnetic spectrum data mining of any one of claims 1 to 7.

Images