Target object description method, device and equipment based on multi-source data fusion

Abstract

The invention discloses a target object description method, device and equipment based on multi-source data fusion, and relates to the technical field of social simulation.The method comprises the steps that a target community needing object description is selected and determined based on an electronic map, an API interface function is called to obtain description data of a target object of the target community in the mobile internet, and the description data are stored in the mobile internet; probability distribution of the target object in each dimension is obtained; obtaining description data of a target object of the target community based on an active investigation mode, and performing conversion processing to obtain probability distribution of the target object in each dimension; and fusing the probability distribution of the target object on each dimension obtained from the mobile internet with the probability distribution of the target object on each dimension obtained in the active investigation mode to obtain a multi-dimensional description of the target object of the target community. According to the method, the community description data is obtained through two modes of mobile internet big data and questionnaire survey, effective complementation is formed, complete crowd portraying of the community can be realized, and community social simulation is effectively supported.

Description

Technical Field

[0001] The present application relates to the field of social simulation technology, and in particular to a target object description method, device and equipment based on multi-source data fusion. Background Art

[0002] With the development of information technology and mobile Internet, people's lifestyles have changed from being single and standardized to being diversified and personalized, showing complex behaviors and spatial needs. Adapting to social development and taking demand as the starting point, how to effectively govern the city, meet people's needs, and optimize resource allocation is a technical problem that needs to be solved urgently. Social simulation is an effective way to solve urban governance problems. By constructing a social environment, modeling entities such as people, cars, roads, and buildings, and simulating the interactions between entities, the operation of society can be effectively simulated and development trends can be predicted. As a basic social unit, the community is a place where residents live their daily lives and has the characteristics of a small society. The simulation of the community can deeply understand the operation of the grassroots society and reflect the general social laws. By simulating natural disasters, disease outbreaks and other events, the effects of different disposal strategies can be analyzed, providing an effective basis for scientific management.

[0003] The basis of community simulation is community population portrait. In order to model community residents, it is necessary to rely on community population portraits to represent and portray people as realistically as possible. At present, most of the research on population portraits focuses on population portraits based on social media, that is, to analyze users based on their posts, comments and other behaviors on social media, to form a population description portrait of social media users, and to achieve the characterization of users' consumption preferences, interests and other aspects. There are many related works that analyze users and their behaviors on social networks. However, social media users are not restricted by region, and it is difficult to filter out users belonging to a certain community for profiling.

[0004] The goal of community simulation is to depict the real people, objects and events in the community. It is necessary to effectively portray the community population to support simulation, that is, to achieve a portrait of the community population. At present, some relevant researchers have proposed a digital portrait method for urban population based on big data. Based on spatiotemporal analysis, urban residents are portrayed from three dimensions of time, space and behavior, and the spatiotemporal behavior characteristics of urban population are analyzed. This portrait only analyzes and portrays the three dimensions related to time and space, focusing on its application in urban planning. It lacks the portrayal of basic attributes of the population such as gender, age, education, consumption, etc., and cannot be directly used for community simulation. Some relevant researchers have also proposed to use the method of spatiotemporal behavior analysis to portray the community population through interview questionnaires. Based on the questionnaire data, four indicators in three aspects of time, space and behavior (single-day rhythm, benchmark geopolarization, activity period, and main behavior patterns) are defined to portray the community population. Nine typical populations are clustered, among which the main behavior patterns are park rest, community leisure, venue culture and sports, and consumption shopping. The main purpose of this community portrait is to clarify the leisure activity needs of community residents and then provide strategic recommendations for the planning of urban public facilities. This portrait also lacks basic attribute information of the population, the sample data volume is small and has corresponding deviations, and it cannot be directly used for community simulation.

[0005] It can be seen that the existing crowd profiling methods are mainly aimed at urban planning and public facilities construction. The dimensions they focus on are only those related to time, space and behavior patterns, and lack basic attribute information of community populations. Therefore, crowd profiling under these dimensions is not enough to support community social simulation. At the same time, the data sources of these methods are single. Some only have a small number of interview questionnaire samples, and some only use LBS (Location Based Services) data, which makes it impossible to collect citizens who do not use mobile phones or use non-smartphones, resulting in sample bias. How to overcome the shortcomings of existing crowd profiling methods and provide crowd profiling results that meet community social simulation has become an urgent problem to be solved. Summary of the invention

[0006] The present application provides a target object description method, device and equipment based on multi-source data fusion, which obtains community characterization data through mobile Internet big data and questionnaire surveys, forming effective complementarity, can achieve a complete population portrait of the community, and effectively support community social simulation.

[0007] In a first aspect, an embodiment of the present application provides a target object description method based on multi-source data fusion, and the target object description method based on multi-source data fusion includes:

[0008] Based on the electronic map, the target community for which object description is required is selected and determined, and the API interface function is called to obtain the description data of the target object in the target community in the mobile Internet, and the probability distribution of the target object in each dimension is obtained;

[0009] Based on active survey, the descriptive data of the target objects in the target community are obtained, and the probability distribution of the target objects in each dimension is obtained through transformation processing;

[0010] The probability distribution of target objects in each dimension obtained from the mobile Internet is integrated with the probability distribution of target objects in each dimension obtained by active survey to obtain a multi-dimensional description of target objects in the target community.

[0011] In combination with the first aspect, in one embodiment, the dimensions include gender, age group, education level, children's age group, life stage, whether or not one has a car, consumption level segment, commuting method, hotel consumption level segment, catering consumption level segment, whether or not one is a frequent traveler, whether or not one is traveling abroad, travel month, travel destination, food type, shopping place, sports and fitness type, entertainment and leisure type, tourist attraction type, smart mobile terminal brand, and smart mobile terminal price.

[0012] In combination with the first aspect, in one implementation, the step of selecting and determining a target community for which object description is required based on an electronic map, and calling an API interface function to obtain description data of a target object in the target community in the mobile Internet, and obtaining a probability distribution of the target object in each dimension, specifically includes:

[0013] Through the electronic web map, select and determine the target community scope for object description;

[0014] Call the API interface function to obtain the descriptive data of the target object in the target community in the mobile Internet within a unit of time, and obtain the probability distribution of the target object in each dimension.

[0015] In combination with the first aspect, in one implementation, the description data of the target objects of the target community is obtained based on the active survey method, and the probability distribution of the target objects in each dimension is obtained by conversion processing, wherein the description data of the target objects of the target community is obtained specifically including:

[0016] Create a questionnaire for collecting descriptive data of each dimension, and actively sample the target objects in the target community in turn during the working day;

[0017] Based on the sampling results, descriptive data of a single target object in the target community is obtained.

[0018] In combination with the first aspect, in one implementation,

[0019] The questionnaire includes multiple dimensions of descriptive data collection for a single target object, and the descriptive results of a single dimension in the questionnaire are single-choice or multiple-choice;

[0020] The single selection is to select one description result in a single dimension for a single target object, and the multiple selection is to select multiple description results in a single dimension for a single target object.

[0021] In combination with the first aspect, in one implementation, the description data of the target object of the target community is obtained based on the active survey method, and the probability distribution of the target object in each dimension is obtained by conversion processing. When the description result of a single dimension is a single choice, the probability distribution of the target object in each dimension is obtained by conversion processing, specifically including:

[0022] Get the number of target objects corresponding to each description result in the current dimension, as well as the total number of target objects;

[0023] Based on the proportion of the number of target objects corresponding to the current description result in the current dimension to the total number of target objects, the probability distribution of the target objects corresponding to each description result in the current dimension is obtained.

[0024] In combination with the first aspect, in one implementation, the description data of the target object of the target community is obtained based on the active survey method, and the probability distribution of the target object in each dimension is obtained by conversion processing. When the description result of a single dimension is multiple choice, the probability distribution of the target object in each dimension is obtained by conversion processing, specifically including:

[0025] Get the number of selected categories of the description results of the current dimension in a single questionnaire, and get the proportion of the current description results of the current dimension in a single questionnaire;

[0026] The proportion of the current description results of the current dimension in all questionnaires is accumulated, and according to the total number of target objects, the probability distribution of the target objects corresponding to the current description results of the current dimension is obtained.

[0027] In combination with the first aspect, in one implementation,

[0028] The probability distribution of the target objects in each dimension obtained from the mobile Internet is fused with the probability distribution of the target objects in each dimension obtained by the active survey method to obtain a multi-dimensional description of the target objects in the target community, wherein the fusion of the probability distribution of the target objects in each dimension obtained by the mobile Internet and the active survey method is achieved based on a linear fusion method;

[0029] The specific implementation of the linear fusion method is:

[0030] For the K probability distributions p to be fused k (k=1,…,K), for each pk Both define a non-negative distance-based score γ k , specifically,

[0031]

[0032] Among them, γ k Represents the probability distribution p k Defined as a non-negative distance-based score, D KL (p k ||p j ) indicates p k and p j KL divergence of

[0033] According to p k and p j For a discrete distribution, we get

[0034]

[0035] Define the weight coefficient, specifically,

[0036]

[0037] Among them, ω k represents the weight coefficient;

[0038] Based on the weight coefficient, the fused probability distribution is obtained.

[0039]

[0040] Among them, p represents the probability distribution after fusion.

[0041] In a second aspect, the present application provides a target object description device based on multi-source data fusion, the target object description device based on multi-source data fusion comprising:

[0042] The first acquisition module is used to select and determine the target community that needs object description based on the electronic map, and call the API interface function to obtain the description data of the target object in the target community in the mobile Internet, and obtain the probability distribution of the target object in each dimension;

[0043] The second acquisition module is used to obtain the description data of the target object of the target community based on the active survey method, and convert it to obtain the probability distribution of the target object in each dimension;

[0044] The fusion module is used to fuse the probability distribution of target objects in various dimensions obtained from the mobile Internet with the probability distribution of target objects in various dimensions obtained by active survey methods to obtain a multi-dimensional description of target objects in the target community.

[0045] In a third aspect, the present application provides a target object description device based on multi-source data fusion, wherein the target object description device based on multi-source data fusion comprises a processor, a memory, and a target object description program based on multi-source data fusion stored in the memory and executable by the processor, wherein when the target object description program based on multi-source data fusion is executed by the processor, the steps of the target object description method based on multi-source data fusion described above are implemented.

[0046] The beneficial effects brought by the technical solution provided in the embodiments of the present application include:

[0047] By obtaining descriptive data of target community target objects in multiple dimensions from multiple sources and fusing the multi-source data after processing, a more complete community population portrait that can support community social simulation is achieved. This application characterizes the community population from multiple attributes such as basic attributes (basic demographic information such as gender and age), wealth attributes, travel preferences, and visit preferences to meet the needs of community social simulation; this application obtains community characterization data through mobile Internet big data and questionnaire surveys to form effective complementarity. After fusion, a complete community population portrait can be achieved, effectively supporting community social simulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a flowchart of a target object description method based on multi-source data fusion in this application;

[0049] Figure 2 This is a functional module diagram of a target object description device based on multi-source data fusion in this application;

[0050] Figure 3 This is a schematic diagram of the hardware structure of the target object description device based on multi-source data fusion in this application. DETAILED DESCRIPTION

[0051] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0052] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0053] In the first aspect, the embodiment of the present application provides a method for describing a target object based on multi-source data fusion. It should be noted that the target object in the present application refers to a community population, and describing the target object is to profile or characterize the community population.

[0054] In one embodiment, referring to Figure 1 , Figure 1 This is a flow chart of the target object description method based on multi-source data fusion in this application. Figure 1 As shown, the target object description method based on multi-source data fusion includes:

[0055] S1: Select and determine the target community for which object description is required based on the electronic map, and call the API interface function to obtain the description data of the target object in the target community in the mobile Internet, and obtain the probability distribution of the target object in each dimension;

[0056] Furthermore, in one embodiment, the dimensions include gender, age group, education level, children's age group, life stage, whether there is a car, consumption level, commuting mode, hotel consumption level, catering consumption level, whether a frequent traveler, whether traveling abroad, travel month, travel destination, food type, shopping place, sports and fitness type, entertainment and leisure type, tourist attraction type, smart mobile terminal brand, smart mobile terminal price. Obtaining the descriptive data of the target object of the target community is to obtain the information of the target object in the above dimensions. For example, for a certain target object, the descriptive data obtained is: gender male, age is between 25 and 30 years old, education level is undergraduate, etc. Of course, for some special cases, such as a certain target object has no children, then in the children's age group dimension, the description data is no children, and the subsequent calculation of the description result in the children's age group dimension is the probability distribution of no children.

[0057] Furthermore, in one embodiment, a target community for which object description is required is selected and determined based on an electronic map, and an API interface function is called to obtain description data of target objects in the target community in the mobile Internet, and the probability distribution of target objects in each dimension is obtained, specifically including:

[0058] S101: Select and determine the target community range for object description through an electronic web map;

[0059] S102: Calling an API (Application Programming Interface) interface function to obtain description data of a target object in a target community in the mobile Internet within a unit time, and obtaining statistically the probability distribution of the target object in each dimension.

[0060] The description data of the target objects of the target community can be obtained through a portal website, an instant messaging operator, etc., based on the relevant publicly available data uploaded or recorded by the target objects of the target community in the mobile Internet.

[0061] Specifically, by delineating the scope of the target community that needs to be described on the electronic web map, calling the API function, the statistical distribution of the community population in the above 21 dimensions in units of days is obtained. For the purpose of privacy protection, some data are interval-based fuzzy processing.

[0062] For a single dimension, it includes multiple description results. For example, the description result of the gender dimension is male or female. If there are 100 target objects in the target community, 59 of them are male and 41 are female, then the probability distribution of the description result of the gender dimension is male is 59 / 100, and the probability distribution of the description result of the gender dimension is female is 41 / 100. The calculation method of the probability distribution of each description result in other dimensions is similar.

[0063] Furthermore, in the actual application process, since the mobile Internet has an automatic statistical function, the probability distribution of each description result of each dimension of the target community target object can be directly obtained through the mobile Internet.

[0064] S2: Obtain the descriptive data of the target objects in the target community through active survey, and convert them into the probability distribution of the target objects in each dimension;

[0065] In order to increase the diversity of data sources, a questionnaire survey was conducted in the community to collect data related to the portrait. Since the data collected by the mobile Internet mainly reflects the information of teenagers, young people and middle-aged groups using smartphones, there is a sample imbalance in the data based on the mobile Internet. The community questionnaire survey was conducted door-to-door on weekdays. On weekdays, teenagers, young people and middle-aged people mostly go out to school or work. The residents at home during the survey were mainly the elderly and children. Therefore, most of the data collected was from the elderly and children, which was also unbalanced. However, the Internet and questionnaire survey data complement each other, and the integration of the two can achieve a more comprehensive portrait of the community population.

[0066] Furthermore, in one embodiment, the description data of the target objects of the target community is obtained based on the active survey method, and the probability distribution of the target objects in each dimension is obtained by conversion processing, wherein the description data of the target objects of the target community is obtained specifically including:

[0067] S201: creating a questionnaire for collecting descriptive data of each dimension, and actively sampling the target objects of the target community in turn during the working day;

[0068] S202: Based on the sampling result, obtain description data of a single target object in the target community.

[0069] The questionnaire includes the collection of descriptive data in multiple dimensions for a single target object, and in the questionnaire, the descriptive result of a single dimension is single choice or multiple choice; single choice means that a single target object selects one descriptive result in a single dimension, and multiple choice means that a single target object selects multiple descriptive results in a single dimension.

[0070] For a single survey object, its descriptive information in the above 21 dimensions needs to be obtained. In the survey questionnaire, each row represents the descriptive data of a target object. The data obtained from the survey is about individuals, and each individual corresponds to a record.

[0071] The characterization of Internet big data is about group distribution, which is not at the same level as the characterization of questionnaire survey data. Questionnaire survey data contains more information and can be converted into distribution characterizations of various dimensions. Therefore, for the descriptive data obtained from the questionnaire survey, it is necessary to convert it into the probability distribution of the target object in each dimension.

[0072] Each question in the questionnaire survey corresponds to a dimension. The answers on some dimensions are single-choice, and the answers on some dimensions are multiple-choice. For example, for the gender dimension, you can only choose male or female, that is, the description result corresponding to the gender dimension is single-choice. For the entertainment and leisure type dimension, you can choose cafes, cinemas, amusement parks, etc. at the same time, that is, the description result corresponding to the entertainment and leisure type dimension is multiple-choice. Of course, a target object only chooses one description result in the entertainment and leisure type dimension, and when performing probability distribution calculation, it is calculated according to the probability distribution calculation method in the case of multiple selection.

[0073] Furthermore, in one embodiment, the description data of the target object of the target community is obtained based on the active survey method, and the probability distribution of the target object in each dimension is obtained by conversion processing. When the description result of a single dimension is a single choice, the probability distribution of the target object in each dimension is obtained by conversion processing, specifically including:

[0074] S211: Obtain the number of target objects corresponding to each description result in the current dimension, and the total number of target objects;

[0075] S212: Based on the proportion of the number of target objects corresponding to the current description result in the current dimension to the total number of target objects, obtain the probability distribution of the target objects corresponding to each description result in the current dimension.

[0076] The following example illustrates the probability distribution calculation method in the single-choice case.

[0077] See Table 1 below, which shows some of the data collected for a questionnaire survey in a community.

[0078] Table 1

[0079] Serial number gender Age (segment) Education Life Stage Age of children 1 female 31-35 years old high school Freelancing 2-5 years old 2 male 31-35 years old junior high school Freelancing 6-12 years old 3 male 40-60 years old junior high school Freelancing other 4 female 25-30 years old Undergraduate Office Workers 2 to 5 years old 5 male Over 61 years old junior high school retire other 6 male 31-35 years old College Office Workers No children 7 female Over 61 years old primary school retire other 8 male 40-45 years old Undergraduate Office Workers 6-12 years old 9 female Over 61 years old primary school retire other 10 female Over 61 years old primary school retire other 11 female Over 61 years old primary school retire other

[0080] Each row in Table 1 represents the information of a resident, and also contains the same 21 dimensions as big data. Obviously, the data obtained from the community questionnaire survey is about individuals, and each individual corresponds to a record.

[0081] As shown in Table 1 above, in the educational dimension, the description results include 7 categories, including primary school, junior high school, high school, junior college, undergraduate, master's degree and doctoral degree. Among the 11 target objects, the number of target objects of each educational level was counted, and 4 people with primary school education, 3 people with junior high school education, 1 person with high school education, 1 person with junior college education, 2 people with undergraduate education, 0 people with master's degree and 0 people with doctoral degree were obtained. Therefore, the probability distribution of the description results of the educational dimension calculated by the questionnaire survey sample is: primary school: 4 / 11=0.3636, junior high school: 3 / 11=0.2727, high school: 0.0909, junior college: 1 / 11=0.0909, undergraduate: 2 / 11=0.1818, master's: 0 / 11=0.0, doctoral: 0 / 11=0.0.

[0082] Furthermore, in one embodiment, based on the active survey method, the description data of the target object of the target community is obtained, and the probability distribution of the target object in each dimension is obtained by conversion processing. When the description result of a single dimension is multiple choice, the probability distribution of the target object in each dimension is obtained by conversion processing, specifically including:

[0083] S221: Obtain the number of selection categories of the description results of the current dimension in a single questionnaire, and obtain the proportion of the current description results of the current dimension in the single questionnaire;

[0084] S222: Accumulate the proportions of the current description results of the current dimension in all questionnaires, and obtain the probability distribution of the target objects corresponding to the current description results of the current dimension according to the total number of target objects.

[0085] The following example illustrates the probability distribution calculation method in the case of multiple selections.

[0086] When the answer is multiple choice, the number of answers is calculated by evenly dividing them. See Table 1 below for an example of dimensions when there are multiple choices.

[0087] Table 2

[0088]

[0089] As shown in Table 2, there are 10 target objects in total. In the shopping place dimension, the description results include 8 categories, including comprehensive shopping malls, digital home appliances, furniture and home building materials, sports and outdoor, books and audio-visual, maternal and child products, jewelry and accessories, and commercial pedestrian streets. The description results of the shopping place dimension of the first target object include comprehensive shopping malls and commercial pedestrian streets, so the comprehensive shopping malls contributed by this data are 1 / 2 people, and the commercial pedestrian streets are 1 / 2 people; the description results of the shopping place dimension of the second target object only include comprehensive shopping malls, so the comprehensive shopping malls contributed by this data are 1 person; the description results of the shopping place dimension of the third target object include comprehensive shopping malls, digital home appliances, and commercial pedestrian streets, so the comprehensive shopping malls contributed by this data are 1 / 3 people, digital home appliances are 1 / 3 people, and commercial pedestrian streets are 1 / 3 people. Other data can be calculated similarly. Add up the number of people who contribute to a certain description result of all data, and then divide it by the total number of people (that is, the number of data items), and you can get the distribution probability of the description result. For example, the number of contributors to shopping malls is 1 / 2+1+1+1 / 3+1+1 / 2+1+1 / 3=5.667 people; the number of contributors to digital home appliances is 1 / 3=0.333 people; the number of contributors to furniture, home furnishings and building materials is 0 people; the number of contributors to sports and outdoor activities is 1 / 3=0.333 people; the number of contributors to books and audio-visual products is 0 people; the number of contributors to maternal and child products is 0 people; the number of contributors to jewelry and accessories is 0 people; the number of contributors to commercial pedestrian streets is 1 / 2+1 / 3+1 / 2+1 / 3+1+1=3.667 people. Divide the number of people describing each result by the total number of people (number of data items) 10, and you will get the probability distribution of the questionnaire data in the shopping dimension as follows:

[0090] Comprehensive shopping malls: 0.5667, digital appliances: 0.0333, furniture and building materials: 0.0, sports and outdoor: 0.0333, books and audio-visual products: 0.0, mother and baby products: 0.0, jewelry and accessories: 0.0, commercial pedestrian street: 0.3667.

[0091] S3: The probability distribution of target objects in each dimension obtained from the mobile Internet is integrated with the probability distribution of target objects in each dimension obtained by active survey to obtain a multi-dimensional description of the target objects in the target community.

[0092] In the present application, the probability distribution of target objects in each dimension obtained from the mobile Internet is fused with the probability distribution of target objects in each dimension obtained by active investigation to obtain a multi-dimensional description of target objects in the target community, wherein the fusion of the probability distribution of target objects in each dimension obtained by mobile Internet and active investigation is achieved based on a linear fusion method.

[0093] According to the above records, mobile Internet data and questionnaire data complement each other, and the fusion of the two can achieve a more comprehensive portrait of the community population. The processing of the questionnaire data in step S2 constructs a probability distribution characterization in each dimension, which is consistent with the characterization level of big data, that is, both use probability distribution to characterize the population in each dimension. Therefore, the fusion of community population portraits comes down to the fusion of probability distributions in various dimensions. Since the two source data are complementary, linear fusion can be used. The key to linear fusion is the determination of the weight coefficient. For mobile Internet big data and questionnaire data, the weight coefficient can be determined based on distance measurement.

[0094] Furthermore, the specific implementation of the linear fusion method in this application is:

[0095] S301: For K probability distributions p to be fused k (k=1,…,K), for each p k Both define a non-negative distance-based score γ k , specifically,

[0096]

[0097] Among them, γ k Represents the probability distribution p k Defined as a non-negative distance-based score, D KL (p k ||p j ) indicates p k and p j KL divergence of

[0098] S302: According to p k and p j For a discrete distribution, we get

[0099]

[0100] S303: Define weight coefficients. Specifically,

[0101]

[0102] Among them, ω k represents the weight coefficient;

[0103] S304: Based on the weight coefficient, obtain the fused probability distribution.

[0104]

[0105] Among them, p represents the probability distribution after fusion.

[0106] For this application, when performing distribution fusion of mobile Internet and questionnaire data portraits, there are two distributions to be fused, K=2, namely the mobile Internet community population portrait and the questionnaire data community population portrait, which are fused separately in each dimension to finally form a fused multi-dimensional community population portrait to achieve a comprehensive population portrait.

[0107] The target object description method based on multi-source data fusion of the embodiment of the present application obtains community population characterization data from multiple sources, including community questionnaire survey data and community population characterization data based on mobile Internet. The data contains information of various dimensions required for social simulation population portraits. The community population characterization data from various sources are then processed separately and unified to the same characterization level. On this basis, the multi-source characterization data is integrated to achieve a more complete community population portrait than single-source data, thereby providing support for community social simulation.

[0108] In a second aspect, an embodiment of the present application also provides a target object description device based on multi-source data fusion.

[0109] In one embodiment, referring to Figure 2 , Figure 2 This is a functional module diagram of the target object description device based on multi-source data fusion in this application. Figure 2 As shown, the target object description device based on multi-source data fusion includes a first acquisition module, a second acquisition module and a fusion module.

[0110] The first acquisition module is used to select and determine the target community that needs object description based on the electronic map, and call the API interface function to obtain the description data of the target object of the target community in the mobile Internet, and obtain the probability distribution of the target object in each dimension; the second acquisition module is used to obtain the description data of the target object of the target community based on the active survey method, and convert it to obtain the probability distribution of the target object in each dimension; the fusion module is used to fuse the probability distribution of the target object in each dimension obtained from the mobile Internet with the probability distribution of the target object in each dimension obtained by the active survey method, and obtain a multi-dimensional description of the target community target object.

[0111] In a third aspect, an embodiment of the present application provides a target object description device based on multi-source data fusion. The target object description device based on multi-source data fusion can be a personal computer (PC), a laptop computer, a server, or other device with data processing capabilities.

[0112] Reference Figure 3 , Figure 3The hardware structure diagram of the target object description device based on multi-source data fusion involved in the embodiment of the present application is shown in FIG. In the embodiment of the present application, the target object description device based on multi-source data fusion may include a processor, a memory, a communication interface and a communication bus.

[0113] The communication bus may be of any type and is used to interconnect the processor, the memory, and the communication interface.

[0114] The communication interface includes input / output (I / O) interface, physical interface and logical interface, etc., which are used to realize the interconnection of devices inside the target object description device based on multi-source data fusion, and the interface used to realize the interconnection between the target object description device based on multi-source data fusion and other devices (such as other computing devices or user devices). The physical interface can be an Ethernet interface, a fiber optic interface, an ATM interface, etc.; the user device can be a display, a keyboard, etc.

[0115] The memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0116] The processor may be a general-purpose processor, and the general-purpose processor may call the target object description program based on multi-source data fusion stored in the memory, and execute the target object description method based on multi-source data fusion provided in the embodiment of the present application. For example, the general-purpose processor may be a central processing unit (CPU). The method executed when the target object description program based on multi-source data fusion is called may refer to the various embodiments of the target object description method based on multi-source data fusion of the present application, which will not be repeated here.

[0117] Those skilled in the art will understand that Figure 3 The hardware structure shown in the figure does not constitute a limitation on the present application, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0118] The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices. The terms "first", "second" and "third" are used to distinguish different objects, etc., and do not represent a sequence, nor do they limit "first", "second" and "third" to different types.

[0119] In the description of the embodiments of the present application, "exemplary", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "for example" or "for example" is intended to present related concepts in a specific way.

[0120] In the description of the embodiments of the present application, unless otherwise specified, “ / ” means or, for example, A / B can mean A or B; the “and / or” in the text is merely a description of the association relationship of associated objects, indicating that three relationships may exist, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, “multiple” refers to two or more than two.

[0121] In some processes described in the embodiments of the present application, multiple operations or steps that appear in a specific order are included, but it should be understood that these operations or steps may not be executed in the order in which they appear in the embodiments of the present application or in parallel, and the sequence number of the operation is only used to distinguish the different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed in sequence or in parallel, and these operations or steps may be combined.

[0122] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, disk, CD) as described above, and includes a number of instructions for a terminal device to execute the methods described in each embodiment of the present application.

[0123] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A target object description method based on multi-source data fusion, It is characterized in that The target object description method based on multi-source data fusion includes: Based on the electronic map, the target community for which object description is required is selected and determined, and the API interface function is called to obtain the description data of the target object in the target community in the mobile Internet, and the probability distribution of the target object in each dimension is obtained; Based on active survey, the descriptive data of the target objects in the target community are obtained, and the probability distribution of the target objects in each dimension is obtained through transformation processing; The probability distribution of target objects in each dimension obtained from the mobile Internet is integrated with the probability distribution of target objects in each dimension obtained by active survey to obtain a multi-dimensional description of target objects in the target community.

2. A target object description method based on multi-source data fusion as claimed in claim 1, Features: The dimensions include gender, age group, education level, children's age group, life stage, whether or not a car is owned, consumption level segment, commuting method, hotel consumption level segment, catering consumption level segment, whether or not a frequent traveler, whether or not traveling abroad, travel month, travel destination, food type, shopping place, sports and fitness type, entertainment and leisure type, tourist attraction type, smart mobile terminal brand, and smart mobile terminal price.

3. A target object description method based on multi-source data fusion as claimed in claim 1, It is characterized in that The method of selecting and determining the target community for which object description is required based on the electronic map, and calling the API interface function to obtain the description data of the target object in the target community in the mobile Internet, and obtaining the probability distribution of the target object in each dimension specifically includes: Through the electronic web map, select and determine the target community scope for object description; Call the API interface function to obtain the descriptive data of the target object in the target community in the mobile Internet within a unit of time, and obtain the probability distribution of the target object in each dimension.

4. A target object description method based on multi-source data fusion as claimed in claim 1, It is characterized in that The description data of the target objects of the target community are obtained based on the active survey method, and the probability distribution of the target objects in each dimension is obtained by conversion processing. The description data of the target objects of the target community are obtained specifically including: Create a questionnaire for collecting descriptive data of each dimension, and actively sample the target objects in the target community in turn during the working day; Based on the sampling results, descriptive data of a single target object in the target community is obtained.

5. A target object description method based on multi-source data fusion as claimed in claim 4, Features: The questionnaire includes multiple dimensions of descriptive data collection for a single target object, and the descriptive results of a single dimension in the questionnaire are single-choice or multiple-choice; The single selection is to select one description result in a single dimension for a single target object, and the multiple selection is to select multiple description results in a single dimension for a single target object.

6. A target object description method based on multi-source data fusion as claimed in claim 5, Features: The description data of the target object of the target community is obtained based on the active survey method, and the probability distribution of the target object in each dimension is obtained by conversion processing. When the description result of a single dimension is a single choice, the probability distribution of the target object in each dimension is obtained by conversion processing, specifically including: Get the number of target objects corresponding to each description result in the current dimension, as well as the total number of target objects; Based on the proportion of the number of target objects corresponding to the current description result in the current dimension to the total number of target objects, the probability distribution of the target objects corresponding to each description result in the current dimension is obtained.

7. A target object description method based on multi-source data fusion as claimed in claim 5, Features: The description data of the target objects of the target community obtained by active investigation is converted to obtain the probability distribution of the target objects in each dimension. When the description result of a single dimension is multiple choice, the probability distribution of the target objects in each dimension is obtained by conversion, specifically including: Get the number of selected categories of the description results of the current dimension in a single questionnaire, and get the proportion of the current description results of the current dimension in a single questionnaire; The proportion of the current description results of the current dimension in all questionnaires is accumulated, and according to the total number of target objects, the probability distribution of the target objects corresponding to the current description results of the current dimension is obtained.

8. A target object description method based on multi-source data fusion as claimed in claim 1, Features: The probability distribution of the target objects in each dimension obtained from the mobile Internet is fused with the probability distribution of the target objects in each dimension obtained by the active survey method to obtain a multi-dimensional description of the target objects in the target community, wherein the fusion of the probability distribution of the target objects in each dimension obtained by the mobile Internet and the active survey method is achieved based on a linear fusion method; The specific implementation of the linear fusion method is: For the K probability distributions p to be fused k (k=1,…,K), for each p k Both define a non-negative distance-based score γ k , specifically, Among them, γ k Represents the probability distribution p k Defined as a non-negative distance-based score, D KL (p k ||p j ) indicates p k and p j KL divergence of According to p k and p j For a discrete distribution, we get Define the weight coefficient, specifically, Among them, ω k represents the weight coefficient; Based on the weight coefficient, the fused probability distribution is obtained. Among them, p represents the probability distribution after fusion.

9. A target object description device based on multi-source data fusion, It is characterized in that The target object description device based on multi-source data fusion includes: The first acquisition module is used to select and determine the target community that needs object description based on the electronic map, and call the API interface function to obtain the description data of the target object in the target community in the mobile Internet, and obtain the probability distribution of the target object in each dimension; The second acquisition module is used to obtain the description data of the target object of the target community based on the active survey method, and convert it to obtain the probability distribution of the target object in each dimension; The fusion module is used to fuse the probability distribution of target objects in various dimensions obtained from the mobile Internet with the probability distribution of target objects in various dimensions obtained by active survey methods to obtain a multi-dimensional description of target objects in the target community.

10. A target object description device based on multi-source data fusion, It is characterized in that The target object description device based on multi-source data fusion includes a processor, a memory, and a target object description program based on multi-source data fusion stored in the memory and executable by the processor, wherein when the target object description program based on multi-source data fusion is executed by the processor, the steps of the target object description method based on multi-source data fusion as described in any one of claims 1 to 8 are implemented.

Images