Method, device and equipment for constructing time-space consumption characteristic knowledge graph of cigarettes

Abstract

The invention relates to the technical field of time-space consumption characteristics of cigarettes, and discloses a method, device and equipment for constructing a cigarette time-space consumption characteristic knowledge graph, and the method comprises the steps: obtaining the cigarette consumption data of a to-be-analyzed geographic region; determining a plurality of space-time grids of the to-be-analyzed geographic area by adopting a space-time grid technology and an internet crawler technology based on the cigarette consumption data; aggregating the total sales volume of cigarettes of each specification in each space-time grid based on the cigarette consumption data; constructing a cigarette consumption characteristic knowledge graph model based on the cigarette consumption data; based on the cigarette consumption characteristic knowledge graph model, determining a recessive relation degree between the total sales volume of each standard cigarette and the cigarette consumption characteristics; a pre-calculation pipeline model is constructed based on the recessive relation degree and the cigarette consumption scenes, and the cigarette space-time consumption characteristic knowledge graph under different space-time combination scenes is determined based on the pre-calculation pipeline model, so that the problem that the cigarette sales situation cannot be accurately judged due to lack of relevance between the cigarette business circle and the space-time is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of spatiotemporal consumption characteristics of cigarettes, and in particular to a method, device and equipment for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes. Background Art

[0002] With the advent of the digital age, all walks of life have accumulated a large amount of production, operation and sales data. How to discover implicit, high-value and continuous information from these data to assist in market judgment, marketing strategy formulation and product design has become an important proposition for enterprises and researchers. In the process of exploring the path of digital transformation, various industries actively use data resources to support the refined development of various activities such as production, operation, research and development, and services, and promote their own transformation and upgrading. For traditional retail industries such as the tobacco industry, it is necessary to understand market consumption trends, competitive situations, future trends, and consumer preferences, and ensure that market information is grasped keenly, timely and comprehensively, promote differentiated cigarette production and personalized cigarette sales, help cigarette marketing activities to be targeted, and marketing decisions to be scientific, and promote the sustainable and healthy development of the cigarette production and sales cycle, so as to obtain new opportunities in the huge market changes.

[0003] In the tobacco industry, market and cigarette researchers usually focus on how to make full use of their own data and external consumer big data, and how to timely grasp the market status and perceive user needs through consumer market insights and market consumption behavior analysis, so as to provide users with more accurate services.

[0004] At present, the market information survey of the tobacco industry mainly adopts manual collection and human experience analysis. Most of them are collected by third-party companies in the form of street survey questionnaires to collect consumer evaluations and opinions, and then rely on statistical methods to form brand analysis reports; because statistics must be based on random sampling to be effective, it is difficult to achieve a truly random survey, so it is impossible to verify the validity and representativeness of the survey results. The existing analysis of the consumption trend of the tobacco industry mostly starts from the source data, uses clustering algorithms and association analysis to mine consumer behavior characteristics, and uses visualization technology to generate cigarette consumption maps to intuitively display the analysis results; this method fails to fully mine the characteristics of the tobacco industry when processing multi-source data, and the visualization analysis in a dynamic interactive environment is also limited to the level of cigarettes and consumers, lacking the correlation interaction analysis with the cigarette business district and time and space, resulting in the inability to accurately judge the cigarette sales trend. How to use multi-source data such as historical cigarette sales data and external business districts to construct a cigarette consumption map, and accurately judge the cigarette sales trend under the support of business fields and regional market characteristics is an urgent problem to be solved. Summary of the invention

[0005] In view of this, the present invention provides a method, device and equipment for constructing a knowledge graph of the temporal and spatial consumption characteristics of cigarettes, so as to solve the problem that the cigarette sales situation cannot be accurately judged due to the lack of interactive analysis of the correlation with the cigarette business circle and time and space.

[0006] In a first aspect, the present invention provides a method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes, the method comprising:

[0007] Obtain cigarette consumption data for the geographic area to be analyzed;

[0008] Based on cigarette consumption data, space-time grid technology and Internet crawler technology are used to determine multiple space-time grids in the geographic area to be analyzed;

[0009] Aggregate the total sales volume of each cigarette specification in each spatiotemporal grid based on cigarette consumption data;

[0010] Construct a knowledge graph model of cigarette consumption characteristics based on cigarette consumption data;

[0011] Determine the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics based on the knowledge graph model of cigarette consumption characteristics;

[0012] A pre-calculated pipeline model is constructed based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenarios. Based on the pre-calculated pipeline model, the knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios is determined.

[0013] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided by the present invention adopts spatiotemporal grid technology to segment multi-source data of cigarette consumption data in time and space dimensions, forming business circles in different time and space, effectively solving the limitations of traditional data analysis in the spatiotemporal dimension, and providing the possibility for in-depth exploration of the implicit association relationship of "people, goods, and places" in the tobacco industry. Based on the cigarette consumption data, the total sales volume of each cigarette specification in each spatiotemporal grid is aggregated, so as to accurately judge the cigarette market situation in different geographical regions and spatial ranges, and understand the target group consumers and store characteristics associated with cigarette products in different cities, different areas, different quarters, different months and other market segments. It provides a decision-making basis for the precise marketing, R&D and production of the cigarette market, and determines the implicit relationship degree of cigarette consumption characteristics based on the knowledge graph model of cigarette consumption characteristics. It builds a pre-calculated pipeline model based on the implicit relationship degree of cigarette consumption characteristics and cigarette consumption scenarios, and determines the knowledge graph of cigarette spatiotemporal consumption characteristics under different spatiotemporal combination scenarios based on the pre-calculated pipeline model. By building a distributed pipeline computing framework, it completes the parallel processing of massive sales business data, ensures the quality and consistency of the data, and then can accurately analyze the cigarette consumption trend, solving the problem of the inability to accurately judge the cigarette sales trend due to the lack of interactive analysis of the correlation with the cigarette business circle and time and space.

[0014] In an optional embodiment, the cigarette consumption data includes the geographical location of the retailer and the cigarette ordering cycle of the retailer;

[0015] Based on cigarette consumption data, the space-time grid technology and Internet crawler technology are used to determine multiple space-time grids for the geographic area to be analyzed, including:

[0016] The geographical area to be analyzed is divided into multiple spatial unit grids using space-time grid technology and Internet crawler technology;

[0017] Obtain the longitude and latitude coordinate information of each spatial unit grid and the longitude and latitude information of the retailer's geographical location;

[0018] Determine the spatial unit grid to which the retailer belongs based on the latitude and longitude coordinate information of each spatial unit grid of the retailer and the latitude and longitude information of the geographical location of the retailer;

[0019] Based on the cigarette ordering cycle of retailers and the spatial unit grids to which the retailers belong, multiple space-time grids are formed; there is a one-to-many relationship between each space-time grid and the retailers.

[0020] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided by the present invention divides the geographic space into a series of grid units through spatiotemporal grid technology and Internet crawler technology, and superimposes the time information of the retailer's cigarette ordering cycle on each grid unit, thereby realizing refined management and analysis of spatiotemporal data of cigarettes, and can accurately judge the cigarette market situation in different geographical regions and spatial ranges, understand the target group consumers and store characteristics associated with cigarette products in different market segments including different cities, different areas, different quarters, and different months, and provide a decision-making basis for precise marketing and R&D and production in the cigarette market.

[0021] In an optional implementation, aggregating the total sales volume of each cigarette specification in each spatiotemporal grid based on cigarette consumption data includes:

[0022] Based on cigarette consumption data, the sales volume of cigarettes of various specifications sold by retailers in each spatiotemporal grid is counted;

[0023] Aggregate the sales volume of various cigarette specifications sold by retailers in each space-time grid to obtain the total sales volume of each cigarette specification in each space-time grid.

[0024] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided by the present invention can timely understand the tobacco market dynamics, grasp market opportunities, and mine characteristic regularity information of different market segments by calculating the total sales volume of each type of cigarette in each spatiotemporal grid.

[0025] In an optional implementation, constructing a knowledge graph model of cigarette consumption characteristics based on cigarette consumption data includes:

[0026] Extracting multiple knowledge entity metadata from cigarette consumption data, the multiple knowledge entity metadata including cigarette attributes, retailers, business districts, business district store characteristics, and business district group consumer characteristics;

[0027] Constructing the first knowledge entity relationship between retailers and cigarette attributes and business districts, the second knowledge entity relationship between business districts and business district store characteristics and business district consumer group characteristics, and the third knowledge entity relationship between cigarette attributes and business district store characteristics and business district consumer group characteristics;

[0028] Construct a knowledge graph model of cigarette consumption characteristics based on the first knowledge entity relationship, the second knowledge entity relationship and the third knowledge entity relationship;

[0029] Among them, the first knowledge entity relationship includes the explicit sales relationship between retailers and cigarette attributes and the inclusion relationship between retailers and business districts; the second knowledge entity relationship includes the inclusion relationship between business districts and business district store characteristics and the inclusion relationship between business districts and business district group consumer characteristics; the third knowledge entity relationship includes the unknown implicit relationship between cigarette attributes and business district store characteristics and the unknown implicit relationship between cigarette attributes and business district group consumer characteristics.

[0030] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided by the present invention constructs a first knowledge entity relationship between retailers and cigarette attributes and business districts, a second knowledge entity relationship between business districts and business district store characteristics and business district group consumer characteristics, and a third knowledge entity relationship between cigarette attributes and business district store characteristics and business district group consumer characteristics; a knowledge graph model of cigarette consumption characteristics is constructed based on the first knowledge entity relationship, the second knowledge entity relationship, and the third knowledge entity relationship, and a knowledge graph covering knowledge type data and their relationships in the field of consumer behavior is formed around entities existing in the field of cigarette consumption and the relationships between entities, so as to deeply understand cigarette consumption behavior, analyze the impact of different types of products on group consumer purchasing behavior, and accurately locate the consumer market and consumer business district of the target product.

[0031] In an optional embodiment, the cigarette consumption characteristics include business district store characteristics and business district group consumer characteristics;

[0032] Based on the knowledge graph model of cigarette consumption characteristics, the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics is determined as follows:

[0033] Based on the preset implicit relationship calculation model, the first implicit relationship value between the total sales volume of each cigarette specification and the characteristics of the business district and store in the preset time and preset regional market in the knowledge graph model of cigarette consumption characteristics and the second implicit relationship value between the total sales volume of each cigarette specification and the characteristics of the consumer group in the business district are calculated;

[0034] Based on the first implicit relationship value and the second implicit relationship value, the first implicit relationship degree between the total sales volume of each cigarette specification and the characteristics of the business district and the second implicit relationship degree between the total sales volume of each cigarette specification and the characteristics of the business district consumers are determined respectively, and the first implicit relationship degree and the second implicit relationship degree are used as the implicit relationship degree between the total sales volume of each cigarette specification and the cigarette consumption characteristics.

[0035] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided by the present invention, based on the explicit relationship that retailers sell cigarettes, business districts contain retailers, and business districts have store and group consumer characteristics, constructs an implicit relationship value calculation model from a specified spatiotemporal range to identify the implicit correlation value of each cigarette with the store characteristics of the business district and the group consumer characteristics of the business district in a specific time and regional market. It can identify the store characteristics of the business district and the group consumer characteristics of the business district that are highly correlated with the sales volume of Huqiu cigarettes, thereby achieving the purpose of analyzing the implicit correlation relationship of "consumers, cigarettes, and market business districts" in the tobacco industry.

[0036] In an optional implementation, the preset implicit relationship calculation model is an implicit relationship degree value calculation model, which is expressed by the following formula:

[0037] R ij =rank(S i1 ,S i2 ,…,S in )

[0038]

[0039]

[0040]

[0041] Among them, R ij represents the sales ranking of the i-th cigarette in the j-th business district, S i1 represents the sales volume of the ith cigarette in the first business district, n represents the number of business districts, and rank represents the ranking function; t im1 represents the average value of the store characteristics and / or consumer characteristics of the mth business district of the i-th cigarette in the top 20% business districts, v imr represents the proportion of the mth label of the i-th cigarette in the rth business district, k represents the number of business districts with the top 20% sales; t im2 represents the average value of the store characteristics of the mth business district and / or the characteristics of the consumer group in the business district in the bottom 80% of the sales districts for the i-th cigarette; p im It represents the relative difference between the label values ​​of the mth business district store characteristics and / or business district consumer group characteristics of the i-th cigarette in the top 20% business districts and the bottom 80% business districts with the highest sales volume, and is used to compare the difference between different labels.

[0042] In an optional embodiment, the cigarette consumption scenario includes multiple cigarette consumption markets and multiple time ranges;

[0043] Based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenarios, a pre-calculated pipeline model is constructed. Based on the pre-calculated pipeline model, the knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios is determined, including:

[0044] Select any one or more cigarette consumption markets from a plurality of cigarette consumption markets and combine them with different time ranges to form different time-space combination scenarios;

[0045] Based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and different spatiotemporal combination scenarios, a pre-calculation pipeline model under different spatiotemporal combination scenarios is constructed;

[0046] Based on the pre-calculated pipeline model, the entities and relationship data between entities in different spatiotemporal combination scenarios are obtained;

[0047] Based on the entities in different spatiotemporal combination scenarios and the relationship data between entities, the knowledge graph of spatiotemporal consumption characteristics of cigarettes in different spatiotemporal combination scenarios is determined.

[0048] The method for constructing the knowledge graph of spatiotemporal consumption characteristics of cigarettes provided by the present invention builds a distributed pipeline computing framework, and forms a process-based configuration of the processing and computing process of dynamic visualization data in different spatiotemporal combination scenarios in the form of components, thereby ensuring the quality and consistency of the data, improving the accuracy and efficiency of data processing, and enhancing the scalability and schedulability of the system. Through the knowledge graph of spatiotemporal consumption characteristics of cigarettes, users can clearly see the features that have a significant impact on cigarette sales, thereby more accurately grasping market dynamics and providing strong data support for market analysis and strategy formulation.

[0049] In an optional implementation, the method for constructing a knowledge graph of cigarette spatiotemporal consumption characteristics further includes:

[0050] The knowledge graph of cigarette spatiotemporal consumption characteristics is visualized, and the cigarette consumption trend is analyzed based on the knowledge graph of cigarette spatiotemporal consumption characteristics in different spatiotemporal combination scenarios.

[0051] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided by the present invention can visualize the knowledge graph of spatiotemporal consumption characteristics of cigarettes, meet the needs of users for rapid positioning and analysis in massive data, help users observe the relationship and structure of data from different perspectives, and enhance the readability of the graph by using visual elements such as node size, color, and thickness of connecting lines, so as to achieve the function of accurately capturing key information and hiding unimportant details, which is conducive to accurate analysis of cigarette consumption trends based on the knowledge graph of spatiotemporal consumption characteristics of cigarettes.

[0052] In a second aspect, the present invention provides a device for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes, the device comprising:

[0053] A data acquisition module, used to acquire cigarette consumption data of a geographic area to be analyzed;

[0054] A space-time grid division module, which is used to determine multiple space-time grids of the geographic area to be analyzed based on cigarette consumption data by using space-time grid technology and Internet crawler technology;

[0055] The sales volume aggregation module is used to aggregate the total sales volume of each cigarette specification in each spatiotemporal grid based on cigarette consumption data;

[0056] A model building module is used to build a knowledge graph model of cigarette consumption characteristics based on cigarette consumption data;

[0057] A relationship calculation module is used to determine the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics based on the knowledge graph model of cigarette consumption characteristics;

[0058] The knowledge graph construction module is used to build a pre-calculated pipeline model based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenarios, and determine the knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios based on the pre-calculated pipeline model.

[0059] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to execute the method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to the first aspect or any corresponding embodiment thereof.

[0060] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 is a flow chart of a method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to an embodiment of the present invention;

[0063] Figure 2 is a flow chart of another method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to an embodiment of the present invention;

[0064] Figure 3 is a flow chart of another method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to an embodiment of the present invention;

[0065] Figure 4 is a diagram of the application of the space-time grid technology according to an embodiment of the present invention;

[0066] Figure 5 is a schematic diagram of a knowledge graph model of cigarette consumption characteristics according to an embodiment of the present invention;

[0067] Figure 6 is a schematic diagram of a pre-computation pipeline model according to an embodiment of the present invention;

[0068] Figure 7 is a schematic diagram of a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to an embodiment of the present invention;

[0069] Figure 8 is a structural block diagram of a device for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to an embodiment of the present invention;

[0070] Fig. 9 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0072] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0073] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, it can also be the internal connection of two components, it can be a wireless connection, or it can be a wired connection. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0074] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0075] The needs for the analysis of the market situation of cigarette consumption include: first, starting from the different regional markets in the target area, making full use of the historical data accumulated in the digital system of the tobacco industry, based on the time series changes of the sales data, visually displaying the cigarette sales situation in the geographical space to which it belongs, and helping to judge the significant changes of cigarettes in different time and space ranges; second, introducing external market data of cigarette consumption, forming the market elements of "people, goods, and places" ("people, goods, and places" correspond to consumers, cigarettes, and market business circles), analyzing the implicit relationship between cigarettes and target consumers and market business circles, and discovering the characteristics of consumer market segments and group consumer characteristics that have a high degree of influence on cigarette sales trends; third, quickly discovering the cigarette sales situation in a specific time, space, crowd, etc., deeply understanding consumers' purchasing habits, preferences and needs, and grasping the hidden laws and trends in cigarette sales data as a whole. Therefore, the embodiment of the present invention provides a method for constructing a knowledge graph of cigarette spatiotemporal consumption characteristics, and constructing a knowledge graph of cigarette spatiotemporal consumption characteristics through multi-source data in cigarette consumption data to achieve the effect of accurately judging the cigarette sales situation.

[0076] According to an embodiment of the present invention, a method embodiment for constructing a knowledge graph of temporal and spatial consumption characteristics of cigarettes is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in an order different from that shown here.

[0077] In this embodiment, a method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes is provided, which can be used in mobile terminals, such as central processing units, servers, etc. Figure 1 is a flow chart of a method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0078] Step S101, obtaining cigarette consumption data of a geographical area to be analyzed.

[0079] Specifically, in order to accurately capture the characteristics of business districts and group consumer characteristics that are highly relevant to various types of cigarettes in the tobacco consumption market, the implicit relationship values ​​between cigarettes and business district markets and business district consumers are mined. This embodiment uses the retailer order data of the tobacco industry production and operation decision-making system, as well as the external data resources introduced, that is, the external data of group consumer population attributes and retailer business district attributes from third-party data service providers, and integrates the basic big data resources for constructing the knowledge graph of spatiotemporal consumption characteristics of cigarettes through external data to enhance internal data. The collected cigarette consumption data includes the cigarette ordering data of market retailers in the geographical areas to be analyzed nationwide, cigarette attribute characteristics (including cigarette specifications, cigarette types), retailer geographic locations, external business district store characteristics data, and business district group consumer characteristics data.

[0080] Step S102, based on the cigarette consumption data, a plurality of space-time grids of the geographic area to be analyzed are determined by using space-time grid technology and Internet crawler technology.

[0081] Specifically, space-time grid technology is a data analysis method that integrates spatial and temporal dimensions. It divides the geographic space into a series of grid cells and superimposes time information on each grid cell, thereby achieving refined management and analysis of space-time data. In space-time grid technology, each grid cell has a clear geographic boundary and time range, making data collection, storage and processing more efficient and accurate. By counting and analyzing grid cells, the distribution characteristics of data in the space-time dimension can be revealed, and the change pattern of data in time and space can be captured and expressed, which helps to understand and deal with complex space-time problems.

[0082] In order to accurately judge the cigarette market situation in different geographical regions and spatial ranges, the space-time grid technology is used to determine multiple space-time grids in the geographical area to be analyzed based on cigarette consumption data. The application diagram of space-time grid technology is shown in the figure below: Figure 4 As shown in the figure, the space-time grid is helpful to analyze the target group consumers and store characteristics associated with cigarette products in different market segments, including different cities, different locations, different quarters, and different months, and provide a decision-making basis for precise marketing and R&D and production in the cigarette market.

[0083] Step S103, aggregating the total sales volume of each cigarette specification in each spatiotemporal grid based on the cigarette consumption data.

[0084] Specifically, since the market demand for cigarette products is affected by holidays, promotional activities, seasonal changes and other factors, in order to timely understand the dynamics of the tobacco market, seize market opportunities, explore the characteristic regularity information of different market segments, and realize dynamic visualization, it is necessary to calculate and process the basic data dimensions at the spatiotemporal grid level. The sales volume of different cigarette products in each spatiotemporal grid is a summary result generated by the aggregation and calculation of a large number of retailer order data. By aggregating the retailer terminal sales data, the retailer-level data is converted into spatiotemporal grids and geographic area-level data associated with different ordering cycles and retailer longitudes and latitudes.

[0085] Step S104, constructing a cigarette consumption characteristic knowledge graph model based on the cigarette consumption data.

[0086] Specifically, in order to reveal the relationship between consumption characteristics in the cigarette market, data related to sales commodities, sales business districts, business district store characteristics, and business district consumer group characteristics within and outside the cigarette enterprise are integrated to support further mining and analysis of implicit relationships in consumption characteristics of cigarette products from a relationship perspective. This embodiment focuses on entities and relationships between entities in the field of cigarette consumption, forms a knowledge graph covering knowledge type data and their relationships in the field of consumer behavior, deeply understands cigarette consumption behavior, analyzes the impact of different types of products on group consumer purchasing behavior, and accurately locates the consumer market and consumer business district of the target product.

[0087] Step S105, determining the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics based on the cigarette consumption characteristics knowledge graph model.

[0088] Specifically, the purpose of determining the implicit relationship degree of cigarette consumption characteristics based on the knowledge graph model of cigarette consumption characteristics is to determine the business district store characteristics and business district consumer group characteristics that are highly correlated with cigarette sales. The implicit correlation values ​​of each cigarette and the business district store characteristics and business district consumer group characteristics under specific time and regional markets can be identified by constructing an implicit relationship value measurement model to determine the business district store characteristics and business district consumer group characteristics that are highly correlated with cigarette sales.

[0089] Step S106, constructing a pre-calculated pipeline model based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenario, and determining the knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios based on the pre-calculated pipeline model.

[0090] Specifically, combined with the scenario requirements of the engineering of the knowledge graph of cigarette sales characteristics, an implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and a pre-calculation pipeline model for cigarette consumption scenarios are established to form pre-calculated entity and relationship data under different spatiotemporal combination scenarios. Based on the pre-calculation pipeline model, the knowledge graph of cigarette spatiotemporal consumption characteristics under different spatiotemporal combination scenarios is determined.

[0091] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided in this embodiment uses spatiotemporal grid technology to segment multi-source data of cigarette consumption data in time and space dimensions to form business districts in different time and space, which effectively solves the limitations of traditional data analysis in the spatiotemporal dimension and provides the possibility for in-depth exploration of the implicit relationship between "people, goods, and places" in the tobacco industry. Based on the cigarette consumption data, the total sales volume of each cigarette specification in each spatiotemporal grid is aggregated, so as to accurately judge the cigarette market situation in different geographical regions and spatial ranges, and understand the target group consumers and stores associated with cigarette products in different market segments including different cities, different areas, different quarters, and different months. Characteristics, provide decision-making basis for precise marketing and R&D and production in the cigarette market, determine the implicit relationship degree of cigarette consumption characteristics based on the knowledge graph model of cigarette consumption characteristics; build a pre-calculated pipeline model based on the implicit relationship degree of cigarette consumption characteristics and cigarette consumption scenarios, and determine the knowledge graph of cigarette spatiotemporal consumption characteristics under different spatiotemporal combination scenarios based on the pre-calculated pipeline model. By building a distributed pipeline computing framework, parallel processing of massive sales business data is completed, ensuring the quality and consistency of the data, and then accurately analyzing the cigarette consumption trend, solving the problem of the inability to accurately judge the cigarette sales trend due to the lack of interactive analysis of the correlation with the cigarette business circle and time and space.

[0092] In this embodiment, a method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes is provided, which can be used in mobile terminals, such as central processing units, servers, etc. Figure 2 is a flow chart of a method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0093] Step S201, obtaining cigarette consumption data of a geographical area to be analyzed.

[0094] In this embodiment, the cigarette ordering data is derived from the retailer order data of the tobacco industry production and operation decision-making system, and includes the fields shown in Table 1. The cigarette attribute characteristics and the retailer geographic location are derived from the internally accumulated retailer basic information data, and include the fields shown in Table 2. The external business district store characteristic data and business district group consumer characteristic data are purchased from third-party operators, companies, and Internet platforms, and include the fields shown in Tables 3 and 4, respectively.

[0095] Table 1 Retailer cigarette ordering data field table

[0096] Serial number Fields type describe 1 retailer_id STRING Retailer ID 2 area STRING area 3 city STRING City 4 cig_name STRING Cigarette Specifications 5 date TIMESTAMP Order Date 6 sls_qty FLOAT Sales

[0097] Table 2 Basic information fields for retailers

[0098]

[0099]

[0100] Table 3 External business district store characteristic data field table

[0101] Serial number Fields type describe 1 business_district_id STRING Business district ID 2 APP_preferences STRING APP Preferences 3 frequent_traveler BOOL Frequent travelers 4 traveling_abroad BOOL Traveling Abroad 5 visitin_preference STRING Visiting Preferences 6 marital_status STRING Marital status 7 travel_distance STRING Travel distance 8 age STRING age 9 life_stage STRING Life Stage 10 availability_of_a_car BOOL Do you have a car? 11 commuting_mode STRING Commuting Mode 12 gender STRING gender 13 education STRING Education 14 occupation STRING Occupational Segmentation 15 distance STRING Distance between work and residence

[0102] Table 4 Data fields of characteristics of consumers in business districts

[0103]

[0104]

[0105] Step S202: Based on the cigarette consumption data, a plurality of space-time grids of the geographic area to be analyzed are determined by using space-time grid technology and Internet crawler technology.

[0106] Specifically, the cigarette consumption data includes the geographical location of the retailer and the cigarette ordering cycle of the retailer. The above step S202 includes:

[0107] Step S2021, using space-time grid technology and Internet crawler technology to divide the geographic area to be analyzed into multiple space unit grids.

[0108] Specifically, this embodiment uses the space-time grid technology, takes the city grid as a unit, and uses the web crawler technology to obtain the maximum and minimum longitude and latitude information of the geographic area market to be analyzed through the public map service API interface, and divides the geographic area market to be analyzed from left to right and from bottom to top to obtain multiple space unit grids of 750m*750m covering the geographic area market, and calculates the maximum and minimum longitude and latitude of each space unit grid. The formula is as follows:

[0109] lat1=min_latitude+row*grid_size*latitude_span

[0110] lon1=min_longitude+col*grid_size*longitude_span

[0111] lat2=lat1+grid_size*latitude_span

[0112] lon2=lon1+grid_size*longitude_span

[0113] Among them, lat1 represents the minimum latitude of the spatial unit grid; lon1 represents the minimum longitude of the spatial unit grid; lat2 represents the maximum latitude of the spatial unit grid; lon2 represents the maximum longitude of the spatial unit grid; min_latitude is the minimum latitude of the geographic area market; min_longitude is the minimum longitude of the geographic area market; row represents the row where the spatial unit grid is located; col represents the column where the spatial unit grid is located; grid_size is the size of the spatial unit grid, that is, 750m; latitude_span represents the latitude span of 1m; longitude_span represents the longitude span of 1m.

[0114] Step S2022, obtaining the longitude and latitude coordinate information of each spatial unit grid and the longitude and latitude information of the retailer's geographic location.

[0115] Specifically, for the longitude and latitude coordinate information of each spatial unit grid, the longitude and latitude coordinates of the four points of the spatial unit grid are obtained through the maximum and minimum longitude and latitude information obtained in step S2021: the lower left corner is (lon1, lat1), the upper left corner is (lon1, lat2), the upper right corner is (lon2, lat2), and the lower right corner is (lon2, lat1).

[0116] The longitude and latitude information of the retailers’ geographic locations can be obtained through the company’s data center. For retailers who lack longitude and latitude information, reverse address resolution can be performed based on the retailers’ geographic locations to obtain longitude and latitude information.

[0117] Step S2023, determining the spatial unit grid to which the retailer belongs based on the latitude and longitude coordinate information of each spatial unit grid of the retailer and the latitude and longitude information of the retailer's geographical location.

[0118] Specifically, the spatial unit grid to which the cigarette retailer belongs is determined based on the specific longitude and latitude information of each spatial unit grid and the longitude and latitude information of the cigarette retailer's geographical location (the minimum grid longitude and latitude <= the retailer's longitude and latitude <= the maximum grid longitude and latitude), and the retailer is mapped to the corresponding spatial unit grid. Each spatial unit grid represents a market business district, realizing a one-to-many relationship between the spatial unit grid and the retailer, and establishing a connection between the retailer and the market business district.

[0119] Step S2024: multiple space-time grids are formed based on the cigarette ordering cycle of the retailer and the spatial unit grid to which the retailer belongs. There is a one-to-many relationship between each space-time grid and the retailer.

[0120] Specifically, in this embodiment, retailer ordering data is used as representative cigarette sales data, and the retailer cigarette ordering cycle "week" is used as the basic unit time period to form a space-time grid with the divided spatial unit grids.

[0121] Step S203, aggregating the total sales volume of each cigarette specification in each spatiotemporal grid based on the cigarette consumption data.

[0122] Specifically, the above step S203 includes:

[0123] Step S2031, based on the cigarette consumption data, statistics are collected on the sales volume of cigarettes of various specifications sold by retailers in each spatiotemporal grid.

[0124] Specifically, the sales volume of each type of cigarettes sold by retailers in each space-time grid is aggregated and calculated.

[0125] Step S2032, aggregating the sales volume of various cigarette specifications sold by retailers in each spatiotemporal grid to obtain the total sales volume of each cigarette specification in each spatiotemporal grid.

[0126] Specifically, the number of cigarettes of each specification sold by retailers in each divided space-time grid is aggregated. For example, the aggregation method of cigarette product A is as follows:

[0127] The total sales volume of cigarette product A within the space-time grid = the sales volume of cigarette product A of retailer 1 + the sales volume of cigarette product A of retailer 2 + … + the sales volume of cigarette product A of retailer N.

[0128] The total sales volume of each type of cigarette in each spatial grid in each ordering cycle is counted in turn, and the time-series ordering data of retailers is matched with the space-time grid.

[0129] Step S204: construct a knowledge graph model of cigarette consumption characteristics based on cigarette consumption data. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0130] Step S205: Determine the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics based on the cigarette consumption characteristics knowledge graph model. Figure 1 Step S105 of the illustrated embodiment will not be described in detail here.

[0131] Step S206: construct a pre-calculated pipeline model based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenario, and determine the knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios based on the pre-calculated pipeline model. Figure 1 Step S106 of the illustrated embodiment will not be described in detail here.

[0132] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided in this embodiment divides the geographic space into a series of grid units and superimposes the time information of the retailer's cigarette ordering cycle on each grid unit, thereby realizing the refined management and analysis of spatiotemporal data of cigarettes, and can accurately judge the cigarette market situation in different geographical regions and spatial ranges, understand the target group consumers and store characteristics associated with cigarette products in different cities, different areas, different quarters, different months and other market segments, and provide a decision-making basis for precise marketing and R&D production in the cigarette market. By calculating the total sales volume of each type of cigarette in each spatiotemporal grid, it is possible to timely understand the dynamics of the tobacco market, grasp market opportunities, and explore the characteristic regularity information of different market segments.

[0133] In this embodiment, a method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes is provided, which can be used in mobile terminals, such as central processing units, servers, etc. Figure 3 is a flow chart of a method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0134] Step S301, obtain cigarette consumption data for the geographic area to be analyzed. Figure 2 Step S201 of the illustrated embodiment will not be described in detail here.

[0135] Step S302: Based on the cigarette consumption data, multiple space-time grids of the geographic area to be analyzed are determined using space-time grid technology and Internet crawler technology. Figure 2 Step S202 of the illustrated embodiment will not be described in detail here.

[0136] Step S303: Aggregate the total sales volume of each cigarette type in each spatiotemporal grid based on cigarette consumption data. Figure 2 Step S203 of the illustrated embodiment will not be described in detail here.

[0137] Step S304: construct a knowledge graph model of cigarette consumption characteristics based on cigarette consumption data.

[0138] Specifically, to construct a knowledge graph in the field of cigarette consumption, it is necessary to extract information from data from different sources based on the constructed knowledge model, identify entities from the data source, and collect specific entity attribute information. This embodiment uses spatiotemporal grid technology as the basis to establish fusion rules to associate business districts and retailers in the cigarette consumption market, and extract cigarette specifications, cigarette types, retailers, business districts, store characteristics, and group consumer characteristics from the associated data table. Using the business district as a carrier, an implicit relationship is established between the characteristics of the business district store, the characteristics of the group consumers in the business district, and the retailers, that is, an implicit relationship; a sales relationship is formed between retailers and the specific cigarette products and different product types sold; through the geographical location of retailers and business districts, an implicit relationship is established between cigarettes and business district stores and group consumers, that is, an implicit relationship.

[0139] Specifically, the above step S304 includes:

[0140] Step S3041, extracting multiple knowledge entity metadata from the cigarette consumption data, the multiple knowledge entity metadata including cigarette attributes, retailers, business districts, business district store characteristics and business district group consumer characteristics.

[0141] Specifically, cigarette attributes include cigarette specifications and cigarette types. In order to build a knowledge graph of cigarette consumption characteristics, entities are extracted from cigarette consumption data in sequence, including cigarette specifications, cigarette types, retailers, business districts, business district store characteristics, and business district consumer group characteristics.

[0142] Step S3042, constructing the first knowledge entity relationship between retailers and cigarette attributes and business districts, the second knowledge entity relationship between business districts and business district store characteristics and business district consumer group characteristics, and the third knowledge entity relationship between cigarette attributes and business district store characteristics and business district consumer group characteristics.

[0143] Specifically, starting from the retailer order data, the internally accumulated retailer basic information data, and the external business district data collected by the third party, the core metadata of each entity and the relationship between the two entities are formulated. The consumption characteristic knowledge entity metadata and knowledge entity relationship are shown in Table 5 and Table 6. The relationship between knowledge entities is constructed in combination with domain knowledge. For example, the relationship between retailers and "cigarette specifications" and "cigarette type" is "sale".

[0144] Table 5 Metadata of consumption characteristic knowledge entity

[0145] Serial number Entity Name Entity Description 1 cig_name Cigarette Specifications 2 cig_type Cigarette type 3 retailer_id Retailer ID 4 business_district_id SUBMARKET 5 business_district_store Features of shopping malls 6 business_district_consumer Consumer characteristics of business district groups

[0146] Table 6 Knowledge entity relationships

[0147]

[0148]

[0149] Step S3043, constructing a knowledge graph model of cigarette consumption characteristics based on the first knowledge entity relationship, the second knowledge entity relationship, and the third knowledge entity relationship;

[0150] Among them, the first knowledge entity relationship includes the explicit sales relationship between retailers and cigarette attributes and the inclusion relationship between retailers and business districts; the second knowledge entity relationship includes the inclusion relationship between business districts and business district store characteristics and the inclusion relationship between business districts and business district group consumer characteristics; the third knowledge entity relationship includes the unknown implicit relationship between cigarette attributes and business district store characteristics and the unknown implicit relationship between cigarette attributes and business district group consumer characteristics.

[0151] Specifically, the knowledge graph model of cigarette consumption characteristics constructed in this embodiment determines the knowledge concepts and scope involved in the knowledge graph based on business logic, including the basic types or classifications of entities, the attributes and attribute value types of entities, and the relationship types between different types of entities. This model adds a new layer of data services based on the knowledge graph structure by describing the attributes and relationships in the cigarette consumption field on the basis of the original consumption field business data resources. The knowledge graph model of the cigarette market consumption characteristics field constructed is as follows: Figure 5 As shown in the figure, the knowledge graph model can realize the analysis of cigarette consumption behavior in the market, understand the consumption level and competition status of different business districts, and link various information such as "consumers, cigarettes, market business districts" involved in cigarette consumption. Different retailers choose to sell different cigarette specifications and cigarette types due to their different scales. Different business districts have different business district store characteristics and business district group consumer characteristics due to regional market resource allocation. Retailers have physical connections with business districts due to their geographical distribution. The above-mentioned various types of data in the field of cigarette consumption will be abstracted from the cleaned system data table, namely the first knowledge entity relationship, the second knowledge entity relationship and the third knowledge entity relationship, and the first knowledge entity relationship, the second knowledge entity relationship and the third knowledge entity relationship will be recombined to form a new data structure. This structure can be described by the three elements of entity, relationship and attribute of the knowledge graph. Based on the model and its description information, the business data in the field of cigarette consumption is mapped to the consumption characteristics knowledge graph model.

[0152] Step S305, determining the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics based on the cigarette consumption characteristics knowledge graph model.

[0153] Specifically, the cigarette consumption characteristics include business district store characteristics and business district consumer group characteristics; the above step S305 includes:

[0154] Step S3051, based on the preset implicit relationship measurement model, calculate the first implicit relationship value between the total sales volume of each cigarette specification and the characteristics of the business district and store at the preset time and preset regional market in the knowledge graph model of cigarette consumption characteristics, and the second implicit relationship value between the total sales volume of each cigarette specification and the characteristics of the business district group consumers.

[0155] Specifically, the preset implicit relationship calculation model is an implicit relationship value calculation model. This embodiment is based on the principle of association analysis in the knowledge graph model of cigarette consumption characteristics. On the basis of the explicit relationship that retailers sell cigarettes, business districts contain retailers, and business districts have store and group consumer characteristics, starting from the specified time and space range, an implicit relationship value calculation model is constructed to identify the implicit correlation value of each cigarette with the store characteristics of the business district and the group consumer characteristics of the business district in a specific time and regional market, with the purpose of finding the store characteristics of the business district and the group consumer characteristics of the business district that are highly correlated with cigarette sales. Regarding the "implicit relationship between cigarette attributes and characteristics of shopping district stores", starting from the idea of ​​"important factors usually account for only a minority, while unimportant factors account for the majority" in the Pareto Principle (also known as the 80 / 20 rule, the Pareto Principle), the main step of the implicit relationship value calculation model is to aggregate the label values ​​of the stores in the hot-selling shopping districts of each cigarette with the label values ​​of the stores in the other non-hot-selling shopping districts, and find out the labels with large differences in stores in the two types of shopping districts as the store characteristics with high relevance to cigarettes; for the "implicit relationship between cigarette attributes and characteristics of shopping district group consumers", the calculation method is the same. Taking the "implicit relationship between cigarette attributes and characteristics of shopping district stores" as an example, the specific calculation logic of the implicit relationship value calculation model is as follows:

[0156] R ij =rank(S i1 ,S i2 ,…,S in )

[0157]

[0158]

[0159]

[0160] Among them, R ij represents the sales ranking of the i-th cigarette in the j-th business district, S i1 represents the sales volume of the ith cigarette in the first business district, n represents the number of business districts, and rank represents the ranking function; t im1 represents the average value of the store characteristics and / or consumer characteristics of the mth business district of the i-th cigarette in the top 20% business districts, v imrrepresents the proportion of the mth label of the i-th cigarette in the rth business district, k represents the number of business districts with the top 20% sales; t im2 represents the average value of the store characteristics of the mth business district and / or the characteristics of the consumer group in the business district in the bottom 80% of the sales districts for the i-th cigarette; p im It represents the relative difference between the label values ​​of the mth business district store characteristics and / or business district consumer group characteristics of the i-th cigarette in the top 20% business districts and the bottom 80% business districts with the highest sales volume, and is used to compare the difference between different labels.

[0161] The preset implicit relationship measurement model can be used to calculate the first implicit relationship value between cigarette attributes and business district store characteristics under the preset time and preset regional market in the knowledge graph model of cigarette consumption characteristics, as well as the second implicit relationship value between cigarette attributes and business district group consumer characteristics.

[0162] Step S3052, based on the first implicit relationship value and the second implicit relationship value, respectively determine the first implicit relationship between the total sales volume of each cigarette specification and the characteristics of the business district and the second implicit relationship between the total sales volume of each cigarette specification and the characteristics of the business district group consumers, and use the first implicit relationship and the second implicit relationship as the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics.

[0163] Specifically, the distribution of cigarette sales usually presents an unbalanced state. The types of cigarettes sold in different business districts are closely related to the economic development level, population density, consumption habits, group consumer characteristics and other factors of the region. The sales volume of the same cigarette in business districts with different consumption levels varies significantly. In order to capture the correlation between cigarettes and the characteristics of the store where they are sold and the characteristics of the group consumer, the Pareto principle is introduced, that is, in any specific group, important factors account for a minority of about 20%, and unimportant factors account for the majority of about 80%. The principle of this principle is applied to the tobacco field to measure the relationship between entities. The first 20% of business districts can be set as hot-selling business districts, and the last 80% of business districts can be set as non-hot-selling business districts. By exploring the differences in the characteristics of cigarettes between hot-selling and non-hot-selling business districts, the store characteristics of the business districts and the characteristics of the group consumer of the business districts with high correlation with cigarettes can be found, that is, the first implicit relationship degree and the second implicit relationship degree can be obtained.

[0164] For example, taking cigarette product A as an example, the implicit relationship between cigarette product A and the "age" attribute value in the consumer characteristics of the business district group is calculated. The attribute values ​​contained in the "age" field are "18-24 years old", "25-30 years old", "31-35 years old", "36-40 years old", "41-45 years old", "46-60 years old", and "over 60 years old". First, the sales volume of cigarette product A in each business district is calculated, and the business districts are ranked according to the sales volume. Based on the Pareto principle and combined with knowledge in the field of cigarette sales, the business districts with the top 20% sales volume are defined as hot-selling business districts, and the business districts with the bottom 80% sales volume are defined as non-hot-selling business districts. Then, the proportion of each characteristic attribute value of "age" in each business district is calculated, and the average value t1 of the proportion of each characteristic attribute value of "age" in the hot-selling business district and the average value t2 of the proportion in the non-hot-selling business district are counted. Finally, the difference between the average values ​​t1 and t2 of "age" is compared, and the characteristic attribute values ​​with large differences are selected as the consumer characteristics of the business district group with a high correlation with cigarette product A in the "age" field.

[0165] The shopping district consumers include 14 features and 207 feature attribute values, such as the "life stage" feature includes "office worker", "middle school student", "college student", "self-employed" and "retired", and the "travel abroad" feature includes "yes" and "no". The shopping district stores include 8 features and 170 feature attribute values, such as the "residential community housing price level" feature includes "low", "medium", "second lowest", "second highest" and "high", and the "residential hotel grade" feature includes "economy chain hotel", "three-star hotel", "four-star hotel", "five-star hotel", "luxury hotel" and "other". According to the above calculation logic, the relationship between each cigarette specification and all feature attribute values ​​is calculated, and the feature attribute values ​​with high correlation with each cigarette are analyzed and mined to accurately locate the characteristics of the target consumption venues and target consumer attributes of cigarettes.

[0166] Step S306, construct a pre-calculated pipeline model based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenario, and determine the knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios based on the pre-calculated pipeline model.

[0167] Specifically, the cigarette consumption scenario includes multiple cigarette consumption markets and multiple time ranges. The above step S306 includes:

[0168] Step S3061: extract any one or more cigarette consumption markets from a plurality of cigarette consumption markets and combine them with different time ranges to form different space-time combination scenarios.

[0169] Specifically, the cigarette consumption market is also called the business district market, and different time ranges can be within a year, a quarter, etc., which are selected according to actual conditions and are not specifically limited here. Any one or more cigarette consumption markets in multiple business district markets are selected and combined with different time ranges to form different time-space combination scenarios.

[0170] Step S3062, based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and different space-time combination scenarios, a pre-calculation pipeline model under different space-time combination scenarios is constructed.

[0171] Specifically, in response to the current problems of different tobacco data collection standards, non-standard requirements, and calculation timeliness when massive marketing business data is aggregated, combined with the scenario requirements of the engineering of the knowledge graph of cigarette sales characteristics, this embodiment uses the pre-calculation and processing technology of the big data pipeline to build a distributed pipeline calculation model, and forms a process-based configuration of the processing and calculation process of dynamic visualization data in different time and space combination scenarios in the form of components. The pipeline description language is formed for the calculation process including the collection of data from different sources, the pre-calculation of time and space grid data, the construction of cigarette consumption characteristic knowledge graph entities and relationship data, and the implicit relationship mining calculation process of cigarette specifications, cigarette types, and cigarette consumption characteristics, so as to realize the integrated full process of collection, calculation, processing, fusion and storage of cigarette consumption characteristic visualization knowledge graph data, support the uninterrupted automatic calculation and update of cigarette sales incremental data over time, and aggregate it into the knowledge graph database.

[0172] The schematic diagram of the pre-calculation pipeline model is as follows Figure 6 As shown in the figure, it includes five parts: pipeline requirement specification, custom development of components, preprocessing of spatiotemporal grid data, entity extraction and automatic calculation of relationships, and pipeline operation process monitoring. The pipeline requirement specification means that the pre-calculation pipeline model needs to conduct detailed data content analysis on the existing internal cigarette data and external procurement data, be familiar with the basic information such as the field, format, characteristics, and completeness of the actual data, as well as the potential associations between different data, and establish corresponding data business tables to store data, including the name, data type, length specification of the table field, etc., to provide preliminary preparation for data preprocessing, grid computing, and knowledge graph association analysis.

[0173] The pre-calculation pipeline model used in this embodiment customizes components according to the needs of building a consumption feature knowledge graph model to achieve engineering of data processing, such as reading retailer order data, processing data, calculating data relationships, etc. By packaging the corresponding functional module code into a program and mounting it to the pipeline system, the components can be used. To ensure that the knowledge graph can make accurate and timely responses according to the specific time, space and area selected by the user, such as Figure 6As shown, this embodiment has built a pipeline component that supports parallel efficient computing and flexible configuration. The component is not only responsible for realizing the grid division of the city, and can automatically associate retailers with the market business district grid, but also can complete the extraction of six major entities in the cigarette consumption market "consumers, cigarettes, and business district markets" and the calculation of implicit relationships between entities, so that when the user selects a specific time and space and area, the purpose of intuitively and quickly displaying the knowledge graph of cigarette consumption characteristics in the selected time and space is achieved. After realizing the engineering construction process of the knowledge graph data and the automatic data update configuration, full-link tracking is implemented for the pre-calculated pipeline model processing process. During the entire pipeline operation process, the smallest independent computing unit (component) is monitored to ensure that the operation and operation of the pipeline are traceable throughout the process.

[0174] Step S3063, obtaining entities and relationship data between entities in different spatiotemporal combination scenarios based on the pre-calculated pipeline model.

[0175] Specifically, entities and relationship data between entities in different spatiotemporal combination scenarios are calculated based on the pre-calculated pipeline model.

[0176] Step S3064, determining the knowledge graph of cigarette spatiotemporal consumption characteristics in different spatiotemporal combination scenarios based on the entities in different spatiotemporal combination scenarios and the relationship data between the entities.

[0177] Specifically, based on the entities and relationship data between entities in different spatiotemporal combination scenarios, a knowledge graph of cigarette spatiotemporal consumption characteristics is obtained, which shows the relationship between cigarette specifications, cigarette types, retailers, business districts and consumer characteristics of business district groups, as well as business district store characteristics.

[0178] Step S307, visually displaying the knowledge graph of cigarette spatiotemporal consumption characteristics and analyzing the cigarette consumption trend based on the knowledge graph of cigarette spatiotemporal consumption characteristics in different spatiotemporal combination scenarios.

[0179] Specifically, from the perspective of visual map layout rendering and interaction, a knowledge map of cigarette spatiotemporal consumption characteristics among cigarette specifications, cigarette types, retailers, business districts and consumer characteristics of business district groups, and business district store characteristics is dynamically displayed, intuitively presenting the implicit relationship between each cigarette and the consumer characteristics of the business district group and the characteristics of the business district store.

[0180] In this embodiment, the pre-calculated pipeline model can provide a visualization engine, which provides functions such as data processing component configuration, data flow configuration and pipeline monitoring in a visual way. Among them, the big data pipeline generator converts the pipeline configured by the visualization method into a pipeline description language, and the Web Service (Web application) calls the RestfulAPI (an interface that meets the REST architecture style) to send the pipeline description language to the execution engine. The execution engine includes a big data pipeline parser, an execution module and a scheduling module. The big data pipeline parser converts the pipeline represented by the pipeline description language into a corresponding directed acyclic graph model. The execution module implements the execution strategy of the pipeline, including stream processing and batch processing. The scheduling submodule schedules the pipeline, and the scheduled pipeline is executed on Yarn (resource scheduling framework). The monitoring module monitors the operation of the pipeline and saves the monitoring data.

[0181] The visualization effect of the knowledge graph of cigarette consumption characteristics is as follows Figure 7 As shown in the figure, the relationship between the top 20 cigarette consumption characteristics in the selected time and space region is displayed in the knowledge graph of cigarette consumption characteristics. In the knowledge graph of cigarette consumption characteristics, entities such as cigarette specifications, business district store characteristics, and business district consumer characteristics are represented by nodes, and the relationship between entities is represented by lines. The size, color, and thickness of the entity nodes in the graph are all of practical significance, which can reflect the sales of different cigarette specifications in the graph, the degree of correlation between business district characteristics and cigarettes, etc., and can also achieve the purpose of analyzing the cigarette consumption trend.

[0182] For cigarette entities, the size of the node reflects the sales volume of the cigarette in the current regional market in the past year. The higher the sales volume of the cigarette, the larger the node in the graph. This facilitates the intuitive presentation of the cigarettes that dominate the market, and facilitates users to quickly identify the best-selling cigarettes. It also provides strong data support for market analysis and strategy formulation. For the characteristics of the store in the business district and the characteristics of the consumer groups in the business district, it is necessary to measure the relationship between the feature label and the cigarettes it is connected to, and then add up all the connection relationship values ​​of each feature in turn, and use the sum of the relationship values ​​to represent the size of the node in the graph; that is, the more cigarettes a business district feature label is associated with, and the higher the correlation value, the larger the node corresponding to the feature will be. Through this design, users can clearly see the characteristics that have a significant impact on cigarette sales, so as to grasp the market dynamics more accurately.

[0183] The relationship between entities is calculated through an implicit relationship measurement model, and the thickness of the connection line is used to display the value of these relationships. By setting the relationship threshold, the graph can be controlled to display only those entities and connections with a relationship value greater than the threshold, thereby avoiding the problem of information overload. In order to realize the association between the graph and time and space, a filter button interaction function is designed. Users can select a specific time and space through the filter button to view the dynamic update of the knowledge graph of cigarette consumption characteristics in the corresponding time and space. This design enables users to more conveniently analyze changes in consumption characteristics in different time and space, providing strong support for the formulation of targeted market strategies.

[0184] The method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes provided in this embodiment breaks the previous method of calculating the association between entities by the Pearson similarity method. It focuses on the need to explore the implicit relationship between entities in the field of tobacco application, and the degree of association should reflect the impact of the store label or consumer label of the business district on the sales of cigarettes. The business district is divided into high-sales business district and low-sales business district according to the proportion method, and the difference in the distribution of label values ​​in the two types of business districts is used to measure the relationship between the business district label and the specific cigarette. By building a distributed pipeline computing framework, the limitations of traditional data processing methods are broken, the standardization and process of data cleaning and fusion processes are realized, and the parallel processing of massive sales business data is completed, ensuring the quality and consistency of data, improving the accuracy and efficiency of data processing, and enhancing the scalability and schedulability of the system. The introduced data-driven visualization framework can meet the needs of users to quickly locate and analyze in massive data, help users observe the relationship and structure of data from different perspectives, and use visual elements such as node size, color, and thickness of connecting lines to enhance the readability of the graph, so as to achieve the function of accurately capturing key information and hiding unimportant details. Combined with the distributed pipeline framework, it can achieve the goal of quickly screening data, filtering data, and building graphs, and realize the engineering processing of each process operation.

[0185] In this embodiment, a device for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes is also provided, and the device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0186] This embodiment provides a device for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes, such as Figure 8 As shown, including:

[0187] The data acquisition module 801 is used to acquire cigarette consumption data of the geographical area to be analyzed.

[0188] The space-time grid division module 802 is used to determine multiple space-time grids of the geographic area to be analyzed based on the cigarette consumption data by using the space-time grid technology and the Internet crawler technology.

[0189] The sales volume aggregation module 803 is used to aggregate the total sales volume of each cigarette specification in each spatiotemporal grid based on the cigarette consumption data.

[0190] The model building module 804 is used to build a cigarette consumption feature knowledge graph model based on cigarette consumption data.

[0191] The relationship calculation module 805 is used to determine the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics based on the cigarette consumption characteristics knowledge graph model.

[0192] The knowledge graph construction module 806 is used to construct a pre-calculated pipeline model based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenario, and determine the knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios based on the pre-calculated pipeline model.

[0193] In some optional implementations, the cigarette consumption data includes the retailer's geographic location and the retailer's cigarette ordering cycle, and the spatiotemporal grid division module 802 includes:

[0194] The grid division unit is used to divide the geographic area to be analyzed into multiple space unit grids by adopting the space-time grid technology and the Internet crawler technology.

[0195] The longitude and latitude information acquisition unit is used to acquire the longitude and latitude coordinate information of each spatial unit grid and the longitude and latitude information of the retailer's geographical location.

[0196] The mapping unit is used to determine the spatial unit grid to which the retailer belongs based on the latitude and longitude coordinate information of each spatial unit grid of the retailer and the latitude and longitude information of the geographical location of the retailer.

[0197] The space-time grid construction unit is used to form multiple space-time grids based on the cigarette ordering cycle of the retailer and the spatial unit grid to which the retailer belongs; there is a one-to-many relationship between each space-time grid and the retailer.

[0198] In some optional implementations, the sales volume aggregation module 803 includes:

[0199] The statistical unit is used to count the sales volume of various cigarette specifications sold by retailers in each space-time grid based on cigarette consumption data.

[0200] The aggregation unit is used to aggregate the sales volume of various cigarette specifications sold by retailers in each space-time grid to obtain the total sales volume of each cigarette specification in each space-time grid.

[0201] In some optional implementations, the model building module 804 includes:

[0202] The metadata extraction unit is used to extract multiple knowledge entity metadata from the cigarette consumption data, and the multiple knowledge entity metadata include cigarette attributes, retailers, business districts, business district store characteristics and business district group consumer characteristics.

[0203] The entity relationship construction unit is used to construct the first knowledge entity relationship between retailers and cigarette attributes and business districts, the second knowledge entity relationship between business districts and business district store characteristics and business district consumer group characteristics, and the third knowledge entity relationship between cigarette attributes and business district store characteristics and business district consumer group characteristics.

[0204] A knowledge graph model building unit, used to build a knowledge graph model of cigarette consumption characteristics based on the first knowledge entity relationship, the second knowledge entity relationship and the third knowledge entity relationship;

[0205] Among them, the first knowledge entity relationship includes the explicit sales relationship between retailers and cigarette attributes and the inclusion relationship between retailers and business districts; the second knowledge entity relationship includes the inclusion relationship between business districts and business district store characteristics and the inclusion relationship between business districts and business district group consumer characteristics; the third knowledge entity relationship includes the unknown implicit relationship between cigarette attributes and business district store characteristics and the unknown implicit relationship between cigarette attributes and business district group consumer characteristics.

[0206] In some optional implementations, the cigarette consumption characteristics include business district store characteristics and business district consumer group characteristics; the relationship calculation module 805 includes:

[0207] The implicit relationship value calculation unit is used to calculate the first implicit relationship value between the total sales volume of each cigarette specification and the characteristics of the business district and stores at a preset time and a preset regional market in the knowledge graph model of cigarette consumption characteristics, and the second implicit relationship value between the total sales volume of each cigarette specification and the characteristics of the consumer group in the business district based on the preset implicit relationship measurement model.

[0208] The relationship determination unit is used to determine the first implicit relationship between the total sales volume of each cigarette specification and the characteristics of the business district and the second implicit relationship between the total sales volume of each cigarette specification and the characteristics of the consumer group in the business district based on the first implicit relationship value and the second implicit relationship value, and use the first implicit relationship and the second implicit relationship as the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics.

[0209] In some optional implementations, the cigarette consumption scenario includes multiple cigarette consumption markets and multiple time ranges, and the knowledge graph construction module 806 includes:

[0210] The combined scenario determination unit is used to extract any one or more cigarette consumption markets from a plurality of cigarette consumption markets and combine them with different time ranges to form different space-time combined scenarios.

[0211] The pre-calculation pipeline model building unit is used to build pre-calculation pipeline models under different space-time combination scenarios based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and different space-time combination scenarios.

[0212] The entity and relationship data acquisition unit between entities is used to acquire the entity and relationship data between entities in different time and space combination scenarios based on the pre-calculated pipeline model.

[0213] The knowledge graph determination unit is used to determine the knowledge graph of cigarette spatiotemporal consumption characteristics in different spatiotemporal combination scenarios based on entities in different spatiotemporal combination scenarios and relationship data between entities.

[0214] In some optional embodiments, the device for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes further includes:

[0215] The visualization module is used to visualize the knowledge graph of cigarette spatiotemporal consumption characteristics and to analyze the cigarette consumption trend based on the knowledge graph of cigarette spatiotemporal consumption characteristics in different spatiotemporal combination scenarios.

[0216] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0217] The device for constructing the knowledge graph of spatiotemporal consumption characteristics of cigarettes in this embodiment is presented in the form of functional units, where the units refer to ASIC (Application Specific Integrated Circuit) circuits, processors and memories that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0218] The embodiment of the present invention also provides a computer device having the above Figure 8 The device shown is for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes.

[0219] See also Fig. 9 , Fig. 9 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Fig. 9As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Fig. 9 A processor 10 is taken as an example.

[0220] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0221] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0222] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0223] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0224] The computer device also includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30 and the output device 40 may be connected via a bus or other means. Fig. 9 The example of connecting through bus is taken in the following.

[0225] The input device 30 can receive input digital or character information, and generate key signal input related to the user settings and function control of the computer device, such as a touch screen, a keypad, a mouse, a track pad, a touch pad, an indicator bar, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (e.g., an LED) and a tactile feedback device (e.g., a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light emitting diode, a display and a plasma display. In some optional embodiments, the display device can be a touch screen.

[0226] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0227] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for constructing a knowledge graph of temporal and spatial consumption characteristics of cigarettes, characterized in that: The method comprises: Obtain cigarette consumption data for the geographic area to be analyzed; Determine multiple space-time grids of the geographic area to be analyzed based on the cigarette consumption data using space-time grid technology and Internet crawler technology; Aggregate the total sales volume of each cigarette specification in each of the spatiotemporal grids based on the cigarette consumption data; Construct a knowledge graph model of cigarette consumption characteristics based on cigarette consumption data; Determine the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics based on the cigarette consumption characteristics knowledge graph model; A pre-calculated pipeline model is constructed based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenarios, and a knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios is determined based on the pre-calculated pipeline model.

2. The method according to claim 1, characterized in that The cigarette consumption data includes the geographical location of the retailer and the cigarette ordering cycle of the retailer; Based on the cigarette consumption data, the multiple space-time grids of the geographic area to be analyzed are determined by using space-time grid technology and Internet crawler technology, including: The geographical area to be analyzed is divided into multiple spatial unit grids using space-time grid technology and Internet crawler technology; Obtain the longitude and latitude coordinate information of each spatial unit grid and the longitude and latitude information of the retailer's geographical location; Determine the spatial unit grid to which the retailer belongs based on the latitude and longitude coordinate information of each spatial unit grid of the retailer and the latitude and longitude information of the geographical location of the retailer; A plurality of space-time grids are formed based on the cigarette ordering cycle of the retailer and the spatial unit grid to which the retailer belongs; and there is a one-to-many relationship between each space-time grid and the retailer.

3. The method according to claim 1, characterized in that Aggregating the total sales volume of each cigarette specification in each spatiotemporal grid based on the cigarette consumption data includes: Based on the cigarette consumption data, the sales volume of cigarettes of various specifications sold by retailers in each spatiotemporal grid is counted; Aggregate the sales volume of various cigarette specifications sold by retailers in each space-time grid to obtain the total sales volume of each cigarette specification in each space-time grid.

4. The method according to claim 1, characterized in that: The construction of a knowledge graph model of cigarette consumption characteristics based on cigarette consumption data includes: Extracting multiple knowledge entity metadata from the cigarette consumption data, wherein the multiple knowledge entity metadata include cigarette attributes, retailers, business districts, business district store characteristics, and business district group consumer characteristics; Constructing first knowledge entity relationships between the retailer and cigarette attributes and business districts, second knowledge entity relationships between business districts and business district store characteristics and business district consumer group characteristics, and third knowledge entity relationships between cigarette attributes and business district store characteristics and business district consumer group characteristics; Constructing a knowledge graph model of cigarette consumption characteristics based on the first knowledge entity relationship, the second knowledge entity relationship and the third knowledge entity relationship; Among them, the first knowledge entity relationship includes the explicit sales relationship between retailers and cigarette attributes and the inclusion relationship between retailers and business districts; the second knowledge entity relationship includes the inclusion relationship between business districts and business district store characteristics and the inclusion relationship between business districts and business district group consumer characteristics; the third knowledge entity relationship includes the unknown implicit relationship between cigarette attributes and business district store characteristics and the unknown implicit relationship between cigarette attributes and business district group consumer characteristics.

5. The method according to claim 1, characterized in that The cigarette consumption characteristics include business district store characteristics and business district consumer group characteristics; The implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics is determined based on the cigarette consumption characteristics knowledge graph model, including: Based on the preset implicit relationship calculation model, the first implicit relationship value between the total sales volume of each cigarette specification and the characteristics of the business district and store in the preset time and preset regional market in the knowledge graph model of cigarette consumption characteristics and the second implicit relationship value between the total sales volume of each cigarette specification and the characteristics of the consumer group in the business district are calculated; Based on the first implicit relationship value and the second implicit relationship value, the first implicit relationship between the total sales volume of each cigarette specification and the characteristics of the business district and the second implicit relationship between the total sales volume of each cigarette specification and the characteristics of the business district consumers are determined respectively, and the first implicit relationship and the second implicit relationship are used as the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics.

6. The method according to claim 5, characterized in that The preset implicit relationship calculation model is an implicit relationship degree calculation model, which is expressed by the following formula: R ij =rank(S i1 ,S i2 ,…,S in ) Among them, R ij represents the sales ranking of the i-th cigarette in the j-th business district, S i1 represents the sales volume of the ith cigarette in the first business district, n represents the number of business districts, and rank represents the ranking function; t im1 represents the average value of the store characteristics and / or consumer characteristics of the mth business district of the i-th cigarette in the top 20% business districts, v imr represents the proportion of the mth label of the i-th cigarette in the rth business district, k represents the number of business districts with the top 20% sales; t im2 represents the average value of the store characteristics of the mth business district and / or the characteristics of the consumer group in the business district in the bottom 80% of the sales districts for the i-th cigarette; p im It represents the relative difference between the label values ​​of the mth business district store characteristics and / or business district consumer group characteristics of the i-th cigarette in the top 20% business districts and the bottom 80% business districts with the highest sales volume, and is used to compare the difference between different labels.

7. The method according to claim 1, characterized in that The cigarette consumption scenarios include multiple cigarette consumption markets and multiple time ranges; A pre-calculation pipeline model is constructed based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and the cigarette consumption scenario, and a knowledge graph of the spatiotemporal consumption characteristics of cigarettes under different spatiotemporal combination scenarios is determined based on the pre-calculation pipeline model, including: Select any one or more cigarette consumption markets from a plurality of cigarette consumption markets and combine them with different time ranges to form different time-space combination scenarios; Based on the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics and different space-time combination scenarios, a pre-calculation pipeline model under different space-time combination scenarios is constructed; Acquire entities and relationship data between entities in different spatiotemporal combination scenarios based on the pre-calculated pipeline model; Based on the entities in the different space-time combination scenarios and the relationship data between the entities, a knowledge graph of the spatiotemporal consumption characteristics of cigarettes in the different space-time combination scenarios is determined.

8. The method according to claim 1, characterized in that The method further comprises: The knowledge graph of the spatiotemporal consumption characteristics of cigarettes is visualized, and the cigarette consumption trend is analyzed based on the knowledge graph of the spatiotemporal consumption characteristics of cigarettes in different spatiotemporal combination scenarios.

9. A device for constructing a knowledge graph of temporal and spatial consumption characteristics of cigarettes, characterized in that: The device comprises: A data acquisition module, used to acquire cigarette consumption data of a geographic area to be analyzed; A space-time grid division module, used to determine a plurality of space-time grids of a geographic area to be analyzed by using space-time grid technology based on the cigarette consumption data; A sales aggregation module, used for aggregating the total sales of each cigarette specification in each of the spatiotemporal grids based on the cigarette consumption data; A model building module is used to build a knowledge graph model of cigarette consumption characteristics based on cigarette consumption data; A relationship calculation module, used to determine the implicit relationship between the total sales volume of each cigarette specification and the cigarette consumption characteristics based on the cigarette consumption characteristics knowledge graph model; The knowledge graph construction module is used to construct a pre-calculated pipeline model based on the implicit relationship between the cigarette consumption characteristics and the cigarette consumption scenarios, and determine the knowledge graph of the cigarette spatiotemporal consumption characteristics under different spatiotemporal combination scenarios based on the pre-calculated pipeline model.

10. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to any one of claims 1 to 8 by executing the computer instructions.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method for constructing a knowledge graph of spatiotemporal consumption characteristics of cigarettes according to any one of claims 1 to 8.

Images