Data processing method and device, computer device, and storage medium

Abstract

Embodiments of the present application disclose a data processing method and device, computer equipment, a storage medium and a computer program product. The data processing method comprises: when a recommendation request for a target industry theme is received, obtaining N target business objects belonging to the target industry theme from M business objects; obtaining a correlation between the target industry theme and each target business object, the correlation between the target industry theme and each target business object being determined according to an industry trend of the target industry theme and a business trend of each target business object respectively; determining a target business object for responding to the recommendation request from the N target business objects according to the N correlations, and outputting the target business object for responding to the recommendation request. The present application can improve the accuracy of content recommendation.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a data processing method, apparatus, computer equipment, storage medium, and computer program product. Background Technology

[0002] In recent years, the rapid development and widespread adoption of internet technology have brought users a wealth of information, satisfying their information needs. However, with the exponential growth of information, it has become difficult for users to sift through this massive amount of data to find the information they truly want. In this context, recommendation systems have emerged to provide precise recommendations, that is, to offer users accurate content and services.

[0003] Currently, recommendation systems filter business content from a vast amount of data based on recent trending events, then push this filtered content to users. However, filtering content solely based on trending events results in a coarse-grained selection process, which reduces the accuracy of content recommendations. Summary of the Invention

[0004] This application provides a data processing method, apparatus, computer equipment, storage medium, and computer program product that can improve the accuracy of content recommendations.

[0005] One embodiment of this application provides a data processing method, including:

[0006] When a recommendation request for a target industry theme is received, N target business objects belonging to the target industry theme are obtained from M business objects; wherein, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, and each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M, and N is a positive integer;

[0007] The correlation between the target industry theme and each target business object is obtained. The correlation between the target industry theme and each target business object is determined based on the industry trend of the target industry theme and the business trend of each target business object.

[0008] Based on N correlations, determine the target business object to respond to the recommendation request from N target business objects, and output the target business object to respond to the recommendation request.

[0009] One embodiment of this application provides a data processing apparatus, including:

[0010] The first acquisition module is used to, when receiving a recommendation request for a target industry theme, acquire N target business objects belonging to the target industry theme from M business objects; wherein, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M, and N is a positive integer;

[0011] The second acquisition module is used to acquire the correlation between the target industry theme and each target business object. The correlation between the target industry theme and each target business object is determined based on the industry trend of the target industry theme and the business trend of each target business object.

[0012] The selection module is used to determine the target business object for responding to the recommendation request from N target business objects based on N relevances, and output the target business object for responding to the recommendation request.

[0013] One aspect of this application provides a computer device, including a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, the processor performs the methods described in the above embodiments.

[0014] One aspect of this application provides a computer storage medium storing a computer program, which includes program instructions. When the program instructions are executed by a processor, they perform the methods described in the above embodiments.

[0015] One embodiment of this application provides a computer program product, which includes a computer program / instruction stored in a computer-readable storage medium. When the computer program / instruction is executed by the processor of a computer device, it performs the methods described in the above embodiments.

[0016] When recommendations need to be made under a specific industry theme, business objects belonging to that industry theme are first screened out. Then, based on the relevance between the business object and the specific industry theme, further business objects to be recommended are selected. This two-layer screening refines the granularity of the screening, improving the accuracy of business object recommendations. Furthermore, this application determines the industry theme to which a business object belongs by using the transaction data volume of the transaction objects it contains and the industry theme of those transaction objects. This ensures the accuracy of the judgment result regarding which industry theme the business object belongs to, further guaranteeing the accuracy of the recommendations. Judging the relevance between the industry theme and the data essence of the business object ensures the accuracy of the relevance results and also improves the accuracy of the business object recommendations. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a system architecture diagram for data processing provided in an embodiment of the present invention;

[0019] Figure 2a This is a schematic diagram of a data processing scenario provided by this application example;

[0020] Figure 2b This is a schematic diagram of a data processing scenario provided by this application example;

[0021] Figure 2c This is a schematic diagram of a data processing scenario provided by this application example;

[0022] Figure 3 This is a schematic diagram of a data processing flow provided in an embodiment of this application;

[0023] Figure 4 This is a schematic diagram of a data processing flow provided in an embodiment of this application;

[0024] Figure 5 This is a schematic diagram of a data processing flow provided in an embodiment of this application;

[0025] Figure 6 This is a schematic diagram of a data processing flow provided in an embodiment of this application;

[0026] Figure 7 This is a system architecture diagram of a blockchain network provided in an embodiment of this application;

[0027] Figure 8 This is an interactive schematic diagram of data processing provided in an embodiment of this application;

[0028] Figure 9 This is a schematic diagram of the structure of a data processing device provided in an embodiment of this application;

[0029] Figure 10 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0030] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0031] Please see Figure 1 This is a system architecture diagram for data processing provided in an embodiment of the present invention. Server 10f establishes a connection with the terminal device cluster through switch 10e and communication bus 10d. The terminal device cluster may include: terminal device 10a, terminal device 10b, ..., terminal device 10c. Taking terminal device 10a as an example, when terminal device 10a receives a recommendation request about a target industry theme, it selects N target business objects belonging to the target industry theme from M business objects. Each of the M business objects contains one or more object transactions and the transaction data volume of the transaction objects. Each transaction object corresponds to an industry theme. The N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M and N is a positive integer. Terminal device 10a obtains the correlation between the target industry theme and each target business object. The correlation between the target industry theme and each business object is determined based on the industry trend of the target industry theme and the business trend of each target business object. Based on the correlation of each target business object, terminal device 10a selects the target business objects to respond to the recommendation request from the N target business objects and outputs the target business objects to respond to the recommendation request.

[0032] The step of determining the target business object for responding to the recommendation request can also be performed by server 10f. Specifically, the terminal device generates a recommendation request for the target industry topic and sends it to server 10f. Server 10f obtains N target business objects and, based on the relevance between the target industry topic and each target business object, selects the target business object for responding to the recommendation request. Server 10f then returns the target business object for responding to the recommendation request to the terminal device.

[0033] Terminal equipment can also be referred to as terminal, user equipment (UE), access terminal, user unit, mobile device, user terminal, wireless communication equipment, user agent, or user device. Terminal equipment can be smart home appliances, handheld devices with wireless communication capabilities (such as smartphones and tablets), computing devices (such as personal computers (PCs), in-vehicle terminals, smart voice interaction devices, wearable devices, or other smart devices, but is not limited to these.

[0034] Server 102 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0035] This application can be applied to various recommendation scenarios (e.g., news recommendation, advertising recommendation, fund recommendation, etc.). First, the business objects under the industry theme are filtered out according to the industry theme. Then, the business objects used for recommendation are selected according to the correlation between the business objects and the industry theme.

[0036] It is understood that in the specific implementation of this application, data related to fund returns, net asset value, price fluctuations, stock price fluctuations, advertising popularity, and the number of reposts, favorites, and comments of text (e.g., news or novels) are involved. When the following embodiments of this application are applied to specific products or technologies, permission or consent from the corresponding company or user is required, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0037] The following are Figures 2a-2c For example, this section explains how to recommend funds to users under specific industry themes.

[0038] Please see Figures 2a-2c This is a schematic diagram of a data processing scenario provided by an example of this application. For example... Figure 2a As shown, Figure 2aThe displayed page is for selecting fund recommendation methods. This page shows four fund recommendation methods: fund ranking recommendation, popular sector recommendation, fund manager recommendation, and index fund recommendation. The fund ranking recommendation method determines the recommended funds based on fund returns; the popular sector recommendation method determines the recommended funds based on the industry to which the fund belongs; the fund manager recommendation method determines the recommended funds based on the fund manager; and the index fund recommendation method determines the recommended funds based on broad-based indices.

[0039] When a user clicks the "Popular Sections" tab, they can choose from four recommendation methods to view popular sections. The terminal device will then be redirected to... Figure 2b The "Hot Books" section page is shown. This page displays multiple industry themes, such as... Figure 2b The current trending topics page displays two industry themes: Smart Home and Chip Concepts. Users can choose their preferred industry theme or a recently trending one. If a user selects the Chip Concepts theme, the device will be redirected to... Figure 2c The fund recommendation page shown.

[0040] exist Figure 2c The corresponding fund recommendation page displays the daily price change and the one-month trend of the "chip concept" industry theme. The trend refers to the percentage change, and the one-month trend refers to the percentage change over the past month. The page also displays two funds recommended to users (Fund A and Fund B) and their relevance to the "chip concept" industry theme. Funds A and B are arranged on the fund recommendation page according to their relevance to the "chip concept" industry theme. Figure 2c As can be seen, the correlation between Fund A and the industry theme "chip concept" is 92.6%, while the correlation between Fund B and the industry theme "chip concept" is 91.6%.

[0041] The specific process by which the terminal device selects Fund A and Fund B from a massive pool of funds is as follows: Since the user selected the chip concept as the industry theme, the terminal device first selects funds belonging to the chip concept from the industry theme fund pool. This pool includes multiple industry themes and funds under each theme. The industry theme a fund belongs to is determined by the number of stocks held by the fund and the industry theme to which those stocks belong. The terminal device then calculates the correlation between the chip concept and each selected fund based on the chip concept's price movement (which could be a one-month, one-year, or six-month trend) and the price movements of the selected funds. The P funds with the highest correlation are selected as the funds to recommend to the user (these funds include Fund A and Fund B).

[0042] The specific process of obtaining N target business objects (such as funds belonging to the chip concept extracted from the industry theme fund pool in the above embodiment), obtaining the correlation between the target industry theme and each target business object (such as the correlation degree in the above embodiment), and determining the target business objects used to respond to the recommendation request (such as fund A and fund B in the above embodiment) can be found in the following. Figures 3-8 Corresponding implementation examples.

[0043] Please see Figure 3 This is a schematic diagram of a data processing procedure provided in an embodiment of this application. This embodiment mainly describes how to determine the funds to be recommended. The data processing may include the following steps:

[0044] Step S101: When a recommendation request for a target industry theme is received, N target business objects belonging to the target industry theme are obtained from M business objects; wherein, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, and each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M, and N is a positive integer.

[0045] Specifically, on the application page of the terminal device (as mentioned above) Figures 2a-2c The corresponding example's hot topic page displays K industry themes (as described above). Figures 2a-2c (Corresponding to the smart home and chip concepts in the embodiments), when any industry topic is selected by the user, the industry topic selected by the user is taken as the target industry topic, and a recommendation request for the target industry topic is generated.

[0046] Optionally, when the terminal device detects that the popularity of a target industry topic exceeds a preset popularity threshold, a recommendation request for the target industry topic is generated. Popularity can be determined based on factors such as netizens' attention, the amount of related information published, the amount of dissemination, and the channels through which the information is disseminated.

[0047] In layman's terms, this means that the terminal device can generate a recommendation request based on current trending industry topics and determine the target business object to respond to the recommendation request. Subsequently, the target business object to respond to the recommendation request can be pushed to the user at a specific time, or directly pushed to the user when the user searches for a target industry topic, thereby reducing user waiting time.

[0048] When a terminal device receives a recommendation request about a target industry topic, it selects from the target industry pool (as described above). Figures 2a-2cIn the corresponding embodiment, N business objects belonging to the target industry theme (referred to as target business objects) are extracted from the industry theme fund pool. N is a positive integer.

[0049] The target business can be funds, and the target industry themes can be: chip industry themes, smart home industry themes, new energy industry themes, etc.

[0050] The object industry pool includes M business objects and the industry theme to which each business object belongs. The industry theme to which a business object belongs is determined by the transaction data volume of the transaction objects it contains and the industry theme to which those transaction objects belong. A business object can belong to multiple different industry themes, and conversely, an industry theme can also contain multiple different business objects, where N ≤ M.

[0051] The trading object can be stocks or other tradable assets. The trading volume of the trading object refers to the number of stocks or the amount of assets. The trading quantity of the trading object can also refer to the proportion of the value of the stocks to the net asset value of the fund or the proportion of the value of the assets to the net asset value of the fund. Each trading object corresponds to an industry theme.

[0052] For example, Company A is a new energy company, and Company B is a chip R&D company. Therefore, the industry theme for Company A's stock (Stock A) is new energy, and the industry theme for Company B's stock (Stock B) is chips. If Fund A holds both Stock A and Stock B, and the value of Stock A accounts for 60% of Fund A's net asset value, while the value of Stock B accounts for 20% of Fund A's net asset value, then in this application, Fund A is the business entity, and this business entity includes two trading entities: Stock A and Stock B. The trading volume of Stock A is 60%, and the trading volume of Stock B is 20%. The industry theme for Stock A is new energy, and the industry theme for Stock B is chips.

[0053] The following explains the specific process of creating an object industry pool:

[0054] The terminal device acquires M business objects, which correspond to K industry themes. The target industry theme is one of the K industry themes. Based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, the terminal device determines the industry theme to which each business object belongs and stores each business object and its industry theme in an object industry pool. For example, if business object A belongs to industry theme A and business object B belongs to industry theme B, then the terminal device can store "business object A - industry theme A" and "business object B - industry theme B" in the object industry pool.

[0055] Since the process of determining the industry theme to which each business object belongs is the same, the following example, using one of the M business objects (referred to as the business object to be processed), illustrates how to determine the industry theme to which the business object to be processed belongs:

[0056] The terminal device divides the transaction objects contained in the pending business object into one or more transaction sets, with the transaction objects in each transaction set belonging to the same industry theme. The terminal device counts the total transaction volume for each transaction set. For any transaction set (called the target transaction set), the transaction data volumes of the transaction objects in the target transaction set are added together to obtain the total transaction volume of the target transaction set. The terminal device identifies the industry theme corresponding to the transaction set whose total transaction volume exceeds a preset transaction volume threshold as the industry theme to which the pending business object belongs.

[0057] For example, the business object to be processed is Fund A. Fund A includes trading object A (trading object A is stock A), trading object B (trading object B is stock B), and trading object C (trading object C is stock C). The trading data volume of stock A is 20%, the trading data volume of stock B is 60%, and the trading data volume of stock C is 20%. The industry theme of stock A is chips, the industry theme of stock B is artificial intelligence, and the industry theme of stock C is artificial intelligence. The terminal device can divide the above three stocks into two trading sets: one trading set 1 includes stock A, and the other trading set 2 includes stock B and stock C. The total trading volume of trading set 1 is 20%, and the total trading volume of trading set 2 is 60% + 20% = 80%. If the trading volume threshold is 50%, then the trading set with a total trading volume greater than the trading volume threshold is trading set 2. Therefore, the industry theme of Fund A is artificial intelligence, which corresponds to the industry theme of trading set 2.

[0058] Step S102: Obtain the correlation between the target industry theme and each target business object. The correlation between the target industry theme and each target business object is determined based on the industry trend of the target industry theme and the business trend of each target business object.

[0059] Specifically, the terminal device acquires the industry trend of the target industry theme and the business trend of each target business object. The industry trend can be the rise and fall of the target industry theme, and the business trend of the target business object can be the rise and fall of a fund. The time period corresponding to the industry trend and the time period corresponding to the business trend can be the same. For example, the industry trend refers to the daily rise and fall of the target industry theme within the period from January 1, 2020 to December 31, 2020, and the business trend refers to the daily rise and fall of the fund within the period from January 1, 2020 to December 31, 2020.

[0060] Calculate the similarity between industry trends and each business trend separately, and use the calculated similarity as the relevance between the target industry theme and each target business object. Since the process of calculating the relevance between the target industry theme and each target business object is the same, the following example uses any one of the N target business objects (referred to as the target business object to be processed) to illustrate how to calculate the relevance between the target industry theme and the target business object to be processed:

[0061] The terminal device calculates the covariance between the industry trend of the target industry theme and the business trend of the target business object to be processed. The formula for calculating the covariance is as follows: Formula (1):

[0062] COV(X,Y)=E(XY)-E(X)E(Y) (1)

[0063] Wherein, COV(X,Y) represents covariance, X is the industry trend of the target industry theme, and Y is the business trend of the target business object to be processed. The industry trend of the target industry theme and the business trend of the target business object to be processed are two vectors with the same dimension.

[0064] The industry standard deviation σ of the industry trend is then calculated separately for each terminal device. X The business standard deviation σ of the business trend of the target business object to be processed Y The correlation between the industry trend of the target industry theme and the business trend of the target business object to be processed is calculated using the following formula (2):

[0065]

[0066] Where P is the correlation between the industry trend of the target industry theme and the business trend of the target business object to be processed.

[0067] Both the industry trend of the target industry theme and the business trend of the target business object are time-related. Therefore, a unit correlation can be calculated for different time periods and assigned different weights. Finally, the weighted sum of the unit correlations of all time periods is the correlation between the industry trend of the target industry theme and the business trend of the target business object. The specific process is as follows:

[0068] There are Q industry trends for the target industry theme, and the number of business trends for the corresponding target business objects to be processed is also Q. Each industry trend and each business trend corresponds to a time period, and the Q time periods of the Q industry trends and the Q time periods of the Q business trends correspond one-to-one. Q is an integer greater than 1.

[0069] For example, between January 1, 2020 and December 31, 2020, 12 time periods (one month per period) can be divided. The daily price fluctuations of the target industry theme in the first month can be combined to form an industry trend, and the daily price fluctuations of the target industry theme in the second month can also be combined to form an industry trend, and so on. Similarly, the daily price fluctuations of a fund in the first month can be combined to form a business trend, and the daily price fluctuations of the fund in the second month can also be combined to form a business trend, and so on.

[0070] Similarly, Q industry trends and Q business trends can be used to calculate Q covariances, as shown in the following formula (3):

[0071] COV t (X t ,Y t )=E(X t Y t )-E(X t )E(Y t (3)

[0072] Among them, COV t (X t ,Y t X represents the covariance over time period t. t Y represents the industry trend of the target industry theme during time period t. t It represents the business trend of the target business object to be processed during time period t.

[0073] Similarly, the terminal device then calculates the industry standard deviation σ of the target industry theme's industry trend during time period t. Xt The business standard deviation σ of the business trend of the target business object to be processed in time period t. Yt The unit correlation between the industry trend of the target industry theme in time period t and the business trend of the target business object to be processed in time period t is calculated using the following formula (4):

[0074]

[0075] Among them, P t It is the unit correlation between the industry trend of the target industry theme in time period t and the business trend of the target business object to be processed in time period t.

[0076] The terminal device obtains the weight of the time period corresponding to each unit of relevance. The time period closer to the current timestamp has a larger weight, and the time period farther from the current timestamp has a smaller weight. The terminal device uses the weights of the Q unit relevances to perform a weighted summation of the Q unit relevances, which yields the relevance between the Q target industry themes and the Q target business objects to be processed. The specific calculation process is as follows: Formula (5)

[0077]

[0078] w t The weight represents the time period t, and P represents the correlation between the Q target industry themes and the Q target business objects to be processed.

[0079] At this point, the terminal device has calculated the correlation between the target industry theme and the target business object to be processed. The correlation between the target industry theme and all target business objects can be calculated in the same way.

[0080] The following explains how to generate Q industry trends and Q business trends for the target business objects to be processed:

[0081] The terminal device acquires multiple unit business trends of the target business object to be processed. Each unit business trend corresponds to a timestamp. Based on the timestamp of each unit business trend, these multiple unit business trends are divided into Q business trends. The timestamps of the unit business trends within the same business trend are in the same time period, and the unit business trends within the same business trend are arranged in ascending order of timestamps. Similarly, the terminal device acquires multiple unit industry trends of the target industry theme. Each unit industry trend corresponds to a timestamp. Based on the timestamp of each unit industry trend, these multiple unit industry trends are divided into Q industry trends. The timestamps of the unit industry trends within the same industry trend are in the same time period, and the unit industry trends within the same industry trend are arranged in ascending order of timestamps.

[0082] For example, there are 6 unit business trends, and the 6 unit business trends are: 0.2, 0.3, 0.5, -0.4, 0.3, -0.3. The timestamps corresponding to these 6 unit business trends are each day between January 1, 2020 and January 6, 2020. There are also 6 unit industry trends, and the 6 unit industry trends are: 0.3, 0.1, -0.5, 0.7, 0.1, -0.2. The timestamps corresponding to these 6 unit business trends are also each day between January 1, 2020 and January 6, 2020. If we divide the time period into 3 days, the above 6 unit business trends can be divided into 2 business trends, namely (0.2, 0.3, 0.5) and (-0.4, 0.3, -0.3). The 6 unit industry trends can also be divided into 2 industry trends, namely (0.3, 0.1, -0.5) and (0.7, 0.1, -0.2).

[0083] Of course, if there is only one industry trend for the target industry theme and only one business trend for the target business object to be processed, then you can directly combine all the unit business trends into a business trend and combine all the unit industry trends into an industry trend.

[0084] Step S103: Based on N correlations, determine the target business object for responding to the recommendation request from the N target business objects, and output the target business object for responding to the recommendation request.

[0085] Specifically, as mentioned above, each target business object has a corresponding correlation. N target business objects will have N correlations. The terminal device can sort these N target business objects in descending order of correlation, and select the top P target business objects as the target business objects to respond to recommendation requests (as described above). Figures 2a-2c (Fund A and Fund B in the corresponding embodiments).

[0086] Optionally, the terminal device may use a target business object with a relevance greater than a preset relevance threshold as the target business object for responding to the recommendation request.

[0087] Optionally, the terminal device determines the business revenue of each target business object based on the business trend of each target business object, adds up the N relevances and N business revenues respectively to obtain the recommendation coefficient of each target business object, and uses the target business objects with recommendation coefficients greater than the recommendation threshold as the target business objects to respond to the recommendation request; or uses the P target business objects with the largest recommendation coefficients as the target business objects to respond to the recommendation request.

[0088] The terminal device displays the target business objects used to respond to recommendation requests on the application page. These target business objects can be arranged on the application page in descending order of relevance.

[0089] Alternatively, when a user searches for a target industry topic, the terminal device displays the target business object on the application page to respond to the recommendation request; or when the current timestamp reaches a preset timestamp, the terminal device displays the target business object on the application page to respond to the recommendation request.

[0090] Please see Figure 4 , Figure 4This is a schematic diagram of a data processing flow provided in an embodiment of this application. First, S industry themes are identified. The method of identifying industry themes is not limited; it can be based on the Shenwan industry theme classification or industry themes identified by the data provider itself. Then, based on the increase in the industry theme index over the past week or month, the terminal device selects K industry themes with the highest increases from the S industry themes and adds them to the industry theme pool. Alternatively, the user can qualitatively select K industry themes to add to the industry theme pool, where K is less than or equal to S.

[0091] Then, the terminal device obtains the holding data of M funds. The fund holding data includes the stock products held by the fund and the proportion of stocks to the fund's net asset value. The fund holding data can be read from the published fund financial reports.

[0092] The terminal device obtains the list of constituent stocks for each industry theme in the industry theme pool from industry theme compilation companies or data providers. In layman's terms, it obtains the industry theme to which each stock belongs.

[0093] For each fund, the terminal device categorizes each stock held by the fund into a corresponding industry theme, and sums up the proportions of each stock to the fund's net asset value to obtain the proportion of the industry theme to the fund's net asset value. This proportion is used as the fund's holding ratio in each industry theme. A threshold is set (the threshold size is unlimited, such as 0.2). When the fund's holding ratio in a certain industry theme exceeds the threshold, the fund is added to the fund candidate pool for that industry theme.

[0094] For each fund, the terminal device acquires market data, which may include one or more of the fund's daily adjusted net asset value and daily price change. This market data is published daily by the fund company. The terminal device also acquires market data for each industry sector, which may include one or more of the industry sector's daily closing price and daily price change.

[0095] The terminal device divides the market data of Fund X in the fund candidate pool of industry theme Y into different time intervals with overlapping market data of industry theme Y, and assigns different weights to them. The correlation coefficient between Fund X and industry theme Y in each time interval is calculated according to the above formula (4). The correlation coefficients of each time interval are weighted and summed using the above formula (5) to obtain the similarity between Fund Y and industry theme X.

[0096] Of course, you can also directly substitute the overlapping market data (including fund market data and industry theme market data) into formula (2) without distinguishing the time interval to obtain the similarity between fund Y and industry theme X.

[0097] For each industry theme, the funds in its candidate fund pool are sorted in descending order of similarity to that industry theme. The top-ranked funds are selected as the funds to choose or recommend, with no limit on the number of funds to select.

[0098] As described above, this application first filters funds based on their holdings in various industry themes, then sorts them according to the similarity of market data between the funds and the industry themes, and finally selects funds. This method is more in line with financial logic and can more accurately select funds related to industry themes, thereby improving the accuracy of fund recommendations.

[0099] Please see Figure 5 , Figure 5 This is a schematic diagram of a data processing procedure provided in an embodiment of this application. This embodiment mainly describes how to determine the advertising data to be recommended. The data processing may include the following steps:

[0100] Step S201: When a recommendation request for a target industry theme is received, N target business objects belonging to the target industry theme are obtained from M business objects; wherein, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, and each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M, and N is a positive integer.

[0101] Specifically, the business object can be advertising data, the transaction object included in the business object can be the promotional object included in the advertising data, the transaction data volume of the transaction object can be the promotional proportion of the promotional object in the advertising data, and the promotional proportion can refer to the pixel ratio of the promotional object in the entire advertising data, or it can be the promotional ratio of the promotional object in the entire advertising data.

[0102] For example, if the advertising data includes celebrity A and snack A, and the advertising share of celebrity A is 0.2 and that of snack A is 0.4, then the advertising targets of this advertising data are celebrity A and snack A. The transaction data volume of the transaction target "celebrity A" is 0.2 and that of the transaction target "snack A" is 0.4. The industry theme of the transaction target "celebrity A" can be the entertainment industry, and the industry theme of the transaction target "snack A" can be the food industry.

[0103] Step S202: Obtain the correlation between the target industry theme and each target business object. The correlation between the target industry theme and each target business object is determined based on the industry trend of the target industry theme and the business trend of each target business object.

[0104] Specifically, the industry trend of a target industry theme can be a combination of the popularity of the target industry theme at multiple points in time. The business trend of a target business entity is a combination of the popularity of target advertising data at multiple points in time. Popularity can be determined based on factors such as netizens' attention, the amount of related information published, the amount of dissemination, and the channels through which the information is disseminated.

[0105] Step S203: Obtain the conversion rate of each target ad data, and sum the N conversion rates and N relevances to obtain the recommendation coefficient of each target ad data.

[0106] Specifically, the terminal device acquires the conversion rate of each target ad data point. The conversion rate refers to the proportion of internet users who click on the ad and enter the promoted website and ultimately convert. The terminal device adds up N conversion rates and N relevance metrics to obtain the recommendation coefficient for each target ad data point.

[0107] Terminal devices can use target ad data with recommendation coefficients greater than the recommendation threshold as target ad data to respond to recommendation requests, or terminal devices can use the P target ad data with the largest recommendation coefficients as target ad data to respond to recommendation requests.

[0108] Step S204: Use the target advertising data with a recommendation coefficient greater than the recommendation threshold as the target advertising data for responding to the recommendation request, and output the target advertising data for responding to the recommendation request.

[0109] Please see Figure 6 , Figure 6 This is a schematic diagram of a data processing procedure provided in an embodiment of this application. This embodiment mainly describes how to determine the text to be recommended. The data processing may include the following steps:

[0110] Step S301: When a recommendation request for a target industry theme is received, N target business objects belonging to the target industry theme are obtained from M business objects; wherein, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, and each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M, and N is a positive integer.

[0111] Specifically, the business object is text, which can be novel text, news text, etc. The transaction object contained in the business object is the keyword group contained in the text, and the transaction data volume of the keyword group is the frequency of the keyword group appearing in the text.

[0112] For example, if the keyword phrases in the text are "electric vehicle" and "market capitalization", and the keyword phrase "electric vehicle" appears 20 times and the keyword phrase "market capitalization" appears 6 times, then the transaction data volume for the trading object "electric vehicle" is 20, the transaction data volume for the trading object "market capitalization" is 6, and the industry theme for the trading object "electric vehicle" is the new energy industry, while the industry theme for the trading object "market capitalization" is the financial industry.

[0113] Step S302: Obtain the correlation between the target industry theme and each target business object. The correlation between the target industry theme and each target business object is determined based on the industry trend of the target industry theme and the business trend of each target business object.

[0114] Specifically, the industry trend of a target industry theme can be a combination of the popularity of the target industry theme at multiple points in time. Similarly, the business trend of a target business object can be a combination of the popularity of the target text at multiple points in time. Popularity can be determined based on factors such as netizens' attention, the amount of related information published, the volume of dissemination, and the channels through which the information is disseminated.

[0115] Step S303: Obtain the attribute information of each target text. The attribute information includes one or more of the following: number of reposts, number of favorites, and number of comments. Combine the attribute information of each target text with the relevance of each target text to form the input features of each target text.

[0116] Step S304: Call the classification model to identify the N input features respectively, obtain the recommendation coefficient of each target text, take the target text with the recommendation coefficient greater than the recommendation threshold as the target text to respond to the recommendation request, and output the target text to respond to the recommendation request.

[0117] Specifically, the terminal device uses a pre-trained classification model to identify the input features of each target text and obtains a recommendation coefficient for each target text.

[0118] The classification model is trained based on a large number of sample texts and their labels. The labels of the sample texts are the recommendation coefficients of the sample texts.

[0119] The terminal device can use target text with a recommendation coefficient greater than the recommendation threshold as the target text to respond to the recommendation request, or the terminal device can use the P target texts with the highest recommendation coefficients as the target text to respond to the recommendation request.

[0120] As described above, applying this application to ad recommendation or text recommendation first filters ad data (or text) belonging to the target industry theme based on the industry theme to which the ad data (or text) belongs. Then, based on the relevance between the ad data (or text) and the target industry theme, ad data (or text) to be recommended is selected. By refining the granularity of the filtering, the ad data to be recommended is guaranteed to belong to the target industry theme and have a high relevance to the target industry theme, thus satisfying the recommendation request and improving recommendation accuracy.

[0121] Please see Figure 7 , Figure 7 This is a system architecture diagram of a blockchain network provided in an embodiment of this application. The terminal device in the foregoing embodiment can be... Figure 7 Node 1, Node 2, Node 3, or Node 4 can all be combined to form a blockchain system. Each node includes a hardware layer, a middleware layer, an operating system layer, and an application layer. From Figure 7 As can be seen, each node in the blockchain system stores the same blockchain data. It is understood that these nodes can include computer devices. The following embodiments are described using a first node and a second node as the execution entities; the first node and the second node are any two nodes among multiple nodes in the blockchain system.

[0122] Please see also Figure 8 , Figure 8 This is an interactive diagram of data processing provided in an embodiment of this application. This embodiment mainly describes the combination of recommendation scenarios and blockchain technology. The data processing includes the following steps S401-S405:

[0123] In step S401, the second node generates a recommendation request for the target industry topic and sends the recommendation request to the first node.

[0124] Specifically, the second node displays K industry topics on the application page. When any industry topic is selected by the user, the second node takes the selected industry topic as the target industry topic and generates a recommendation request for the target industry topic.

[0125] Optionally, when the second node detects that the current popularity of a target industry topic exceeds a preset popularity threshold, the second node generates a recommendation request for the target industry topic. Popularity can be determined based on factors such as netizens' attention, the amount of related information published, the amount of dissemination, and the channels through which the information is disseminated.

[0126] In layman's terms, the second node can generate recommendation requests based on current trending industry topics and determine the target business object to respond to those requests. Subsequently, the target business object can be pushed to the user at a specific time to respond to the recommendation request, or the second node can directly push the target business object to the user when they search for a target industry topic.

[0127] Step S402: The first node obtains the block height corresponding to the target industry theme, extracts the target block corresponding to the block height from the blockchain, and reads N target business objects belonging to the target industry theme from the target block; wherein, the blockchain stores M business objects, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, N≤M, where N is a positive integer.

[0128] Specifically, a blockchain consists of multiple blocks, which store M business objects. These M business objects correspond to K industry themes, and each block stores only one business object under an industry theme. That is, business objects in the same block belong to the same industry theme.

[0129] The first node searches for the block height corresponding to the target industry theme in the blockchain industry theme record table, extracts the block corresponding to that block height from the blockchain (called the target block), and reads N target business objects belonging to the target industry theme from the block body of the target block.

[0130] The blockchain industry theme record table records multiple industry themes, as well as the block height of each industry theme.

[0131] The specific process of generating a blockchain is as follows:

[0132] The first node obtains M business objects. Based on the transaction data volume of each business object and the corresponding industry theme, it determines the industry theme to which each business object belongs. Business objects belonging to the same industry theme are grouped into a business object heap, and each business object heap is stored in a block (one business object heap is stored in one block) and added to the blockchain. The first node generates an industry theme record table based on the industry theme corresponding to each block and the block height.

[0133] The specific process for determining the industry theme to which each business object belongs can be found above. Figure 3 The corresponding step S101 in the embodiment.

[0134] Step S403: The first node obtains the correlation between the target industry theme and each target business object. The correlation between the target industry theme and each target business object is determined based on the industry trend of the target industry theme and the business trend of each target business object.

[0135] Step S404: The first node determines the target business object to respond to the recommendation request from N target business objects based on N relevances.

[0136] The specific process of step S404 can be found above. Figure 3 The corresponding step S103 in the embodiment.

[0137] In step S405, the first node sends the target business object used to respond to the recommendation request to the second node.

[0138] As mentioned above, relying on the complete and immutable properties of blockchain, it can be guaranteed that the recommendation request is trustworthy and has not been tampered with. Furthermore, the business object and the industry theme to which the business object belongs are stored on the blockchain. Therefore, the target business object used to respond to the recommendation request is trustworthy, which can guarantee the security and accuracy of the recommendation process.

[0139] For further details, please see Figure 9 This is a schematic diagram of the structure of a data processing device provided in an embodiment of this application. Figure 9 As shown, the data processing device 1 can be applied to the above-mentioned Figures 3-8 The corresponding terminal device or first node in the embodiment. Specifically, the data processing device 1 may be a computer program (including program code) running on a computer device, for example, the data processing device 1 is an application software; the data processing device 1 may be used to execute the corresponding steps in the method provided in the embodiments of this application.

[0140] The data processing device 1 may include: a first acquisition module 11, a second acquisition module 12, and a selection module 13.

[0141] The first acquisition module 11 is used to acquire N target business objects belonging to the target industry theme from M business objects when a recommendation request for the target industry theme is received; wherein, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M and N is a positive integer;

[0142] The second acquisition module 12 is used to acquire the correlation between the target industry theme and each target business object. The correlation between the target industry theme and each target business object is determined based on the industry trend of the target industry theme and the business trend of each target business object.

[0143] Selection module 13 is used to determine the target business object for responding to the recommendation request from N target business objects based on N relevances, and output the target business object for responding to the recommendation request.

[0144] In one possible implementation, the M business objects contain transaction objects corresponding to K industry themes, and the target industry theme is one of the K industry themes.

[0145] The device 1 further includes a third acquisition module 14.

[0146] The third acquisition module 14 is used to acquire M business objects, determine the industry theme to which each business object belongs based on the amount of transaction data contained in each business object and the industry theme corresponding to the transaction object, and associate and store each business object and the industry theme to which each business object belongs to the object industry pool.

[0147] When the first acquisition module 11 is used to acquire N target business objects belonging to the target industry theme from M business objects, it is specifically used for:

[0148] Extract N target business objects belonging to the target industry theme from the object industry pool.

[0149] In one possible implementation, the business object to be processed is one of M business objects, and the third acquisition module 14, when determining the industry theme to which the business object to be processed belongs, is specifically used for:

[0150] The transaction objects contained in the business objects to be processed are divided into one or more transaction sets, and the transaction objects in each transaction set belong to the same industry theme;

[0151] The total transaction volume of each transaction set is calculated. The total transaction volume of the target transaction set is obtained by summing up the transaction data of the transaction objects in the target transaction set. The target transaction set is one of one or more transaction sets.

[0152] The industry theme corresponding to the transaction set whose total transaction volume exceeds the transaction volume threshold is taken as the industry theme to which the business object to be processed belongs.

[0153] In one possible implementation, the target business object to be processed is one of N target business objects. When the second acquisition module 12 is used to acquire the correlation between the target industry theme and the target business object to be processed, it is specifically used for:

[0154] Calculate the covariance between the industry trend of the target industry theme and the business trend of the target business object to be processed, and calculate the industry standard deviation of the industry trend of the target industry theme and the business standard deviation of the business trend of the target business object to be processed.

[0155] The correlation between the target industry theme and the target business object to be processed is determined based on the covariance, the industry standard deviation, and the business standard deviation.

[0156] In one possible implementation, the number of industry trends of the target industry theme and the number of business trends of the target business objects to be processed are both Q. Each industry trend and each business trend corresponds to a time period, and the Q time periods of the Q industry trends and the Q time periods of the Q business trends correspond one-to-one. Q is an integer greater than 1.

[0157] When the second acquisition module 12 is used to determine the correlation between the target industry theme and the target business object to be processed based on the covariance, the industry standard deviation, and the business standard deviation, it is specifically used for:

[0158] The unit correlation between the target industry theme and the target business object to be processed over Q time periods is determined based on the covariance, the industry standard deviation, and the business standard deviation.

[0159] The weights are determined based on the time periods corresponding to the relevance of each unit.

[0160] The Q-unit correlations are weighted and superimposed to form the correlation between the target industry theme and the target business object to be processed.

[0161] In one possible implementation, the device 1 further includes a combination module 15.

[0162] The combination module 15 is used to obtain multiple unit business trends of the target business object to be processed. Each unit business trend has a corresponding timestamp. Based on the timestamp corresponding to each unit business trend, the multiple unit business trends are divided into Q business trends. The timestamps of the unit business trends in the same business trend are in the same time period, and the unit business trends in the same business trend are arranged in ascending order of timestamp.

[0163] The combination module 15 is also used to obtain multiple unit industry trends of the target industry theme. Each unit industry trend has a corresponding timestamp. Based on the timestamp corresponding to each unit industry trend, the multiple unit industry trends are divided into Q industry trends. The timestamps of the unit industry trends in the same industry trend are in the same time period, and the unit industry trends in the same industry trend are arranged in ascending order of timestamp.

[0164] In one possible implementation, when the selection module 13 determines the target business object for responding to the recommendation request from N target business objects based on N relevances, it is specifically used for:

[0165] Sort the N target business objects in descending order of relevance, and select the first P target business objects as the target business objects to respond to the recommendation request, where P is a positive integer; or...

[0166] Target business objects with a relevance greater than a relevance threshold are selected as the target business objects for responding to the recommendation request; or...

[0167] Based on the business trend of each target business object, the business revenue of each target business object is determined. The N relevances and N business revenues are superimposed to obtain the recommendation coefficient of each target business object. The target business objects with recommendation coefficients greater than the recommendation threshold are used as the target business objects to respond to the recommendation request.

[0168] In one possible implementation, the device 1 further includes a generation module 16.

[0169] The generation module 16 is used to display K industry topics on the application page, and generate a recommendation request for the target industry topic when the target industry topic is selected; or, it is used to generate a recommendation request for the target industry topic when the popularity of the target industry topic is greater than the popularity threshold.

[0170] When the selection module 13 is used to output the target business object for responding to the recommendation request, it is specifically used for:

[0171] The application page displays the target business object used to respond to the recommendation request.

[0172] In one possible implementation, the target business object is the target advertising data, the transaction object contained in the target business object is the promotional object contained in the target advertising data, and the transaction data volume of the promotional object is the promotional proportion of the promotional object in the advertising data.

[0173] When the selection module 13 determines the target business object for responding to the recommendation request from N target business objects based on N relevances, it is specifically used for:

[0174] Obtain the conversion rate for each target ad data point, and then sum the N conversion rates and N relevances to obtain the recommendation coefficient for each target ad data point.

[0175] The target advertising data with a recommendation coefficient greater than the recommendation threshold will be used as the target advertising data to respond to the recommendation request.

[0176] In one possible implementation, the target business object is the target text, and the transaction object included in the target business object is the keyword group of the target text. The transaction data volume of the keyword group is the frequency of occurrence of the keyword group in the text.

[0177] When the selection module 13 determines the target business object for responding to the recommendation request from N target business objects based on N relevances, it is specifically used for:

[0178] Obtain the attribute information for each target text, including one or more of the following: number of reposts, number of favorites, and number of comments;

[0179] The attribute information of each target text is combined with the relevance of each target text to form the input features of each target text;

[0180] The classification model is called to identify the N input features respectively, and the recommendation coefficient of each target text is obtained;

[0181] Target text with a recommendation coefficient greater than the recommendation threshold will be used as the target text to respond to the recommendation request.

[0182] In one possible implementation, the device 1 is installed on the first node for execution, and the recommendation request is generated by the second node and sent to the first node;

[0183] When the first acquisition module 11 is used to acquire N target business objects belonging to the target industry theme from M business objects, it is specifically used for:

[0184] Obtain the block height corresponding to the target industry theme, extract the target block corresponding to the block height from the blockchain, and read N target business objects belonging to the target industry theme from the target block. The blockchain stores M business objects.

[0185] The selection module is used to output the target business object for responding to the recommendation request, specifically by sending the target business object for responding to the recommendation request to the second node.

[0186] According to an embodiment of the present invention, Figures 3-8 Each step involved in the method shown can be performed by... Figure 9The data processing is performed by the various modules within the shown data processing device. For example, Figure 3 Steps S101-S103 shown Figure 5 Steps S201-S204 are shown. Figure 6 Steps S301-S304 shown in the figure Figure 8 Steps S402-S404 shown can be respectively made by Figure 9 The first acquisition module 11, the second acquisition module 12, the selection module 13, the third acquisition module 14, the combination module 15, and the generation module 16 shown are used to execute this.

[0187] Further, please see Figure 10 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. The above... Figures 3-8 In the corresponding embodiment, the terminal device or the first node can be a computer device 1000. For example... Figure 10 As shown, computer device 1000 may include a user interface 1002, a processor 1004, an encoder 1006, and a memory 1008. A signal receiver 1016 is used to receive or transmit data via a cellular interface 1010, a Wi-Fi interface 1012, ..., or an NFC interface 1014. The encoder 1006 encodes the received data into a data format that can be processed by a computer. The memory 1008 stores a computer program, and the processor 1004 is configured to execute the steps in any of the above method embodiments via the computer program. The memory 1008 may include volatile memory (e.g., dynamic random access memory DRAM) and may also include non-volatile memory (e.g., one-time programmable read-only memory OTPROM). In some instances, the memory 1008 may further include memory remotely located relative to the processor 1004, which can be connected to the computer device 1000 via a network. The user interface 1002 may include a keyboard 1018 and a display 1020.

[0188] exist Figure 10 In the computer device 1000 shown, the processor 1004 can be used to call computer programs stored in the memory 1008 to achieve:

[0189] When a recommendation request for a target industry theme is received, N target business objects belonging to the target industry theme are obtained from M business objects; wherein, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, and each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M, and N is a positive integer;

[0190] The correlation between the target industry theme and each target business object is obtained. The correlation between the target industry theme and each target business object is determined based on the industry trend of the target industry theme and the business trend of each target business object.

[0191] Based on N correlations, determine the target business object to respond to the recommendation request from N target business objects, and output the target business object to respond to the recommendation request.

[0192] In one embodiment, the transaction objects contained in the M business objects correspond to K industry themes, and the target industry theme is one of the K industry themes;

[0193] Processor 1004 also performs the following steps:

[0194] Get M business objects;

[0195] Based on the transaction data volume of each business object and the industry theme corresponding to the transaction object, determine the industry theme to which each business object belongs, and associate and store each business object and the industry theme to which each business object belongs in the object industry pool.

[0196] When the processor 1004 retrieves N target business objects belonging to the target industry theme from M business objects, it specifically performs the following steps:

[0197] Extract N target business objects belonging to the target industry theme from the object industry pool.

[0198] In one embodiment, the business object to be processed is one of M business objects. When the processor 1004 determines the industry theme to which the business object to be processed belongs, it specifically performs the following steps:

[0199] The transaction objects contained in the business objects to be processed are divided into one or more transaction sets, and the transaction objects in each transaction set belong to the same industry theme;

[0200] The total transaction volume of each transaction set is calculated. The total transaction volume of the target transaction set is obtained by summing up the transaction data of the transaction objects in the target transaction set. The target transaction set is one of one or more transaction sets.

[0201] The industry theme corresponding to the transaction set whose total transaction volume exceeds the transaction volume threshold is taken as the industry theme to which the business object to be processed belongs.

[0202] In one embodiment, the target business object to be processed is one of N target business objects. When the processor 1004 executes the process of obtaining the correlation between the target industry topic and the target business object to be processed, it specifically performs the following steps:

[0203] Calculate the covariance between the industry trend of the target industry theme and the business trend of the target business object to be processed, and calculate the industry standard deviation of the industry trend of the target industry theme and the business standard deviation of the business trend of the target business object to be processed.

[0204] The correlation between the target industry theme and the target business object to be processed is determined based on the covariance, the industry standard deviation, and the business standard deviation.

[0205] In one embodiment, the number of industry trends of the target industry theme and the number of business trends of the target business object to be processed are both Q. Each industry trend and each business trend corresponds to a time period, and the Q time periods of the Q industry trends and the Q time periods of the Q business trends correspond one-to-one. Q is an integer greater than 1.

[0206] When the processor 1004 determines the correlation between the target industry theme and the target business object to be processed based on the covariance, the industry standard deviation, and the business standard deviation, it specifically performs the following steps:

[0207] The unit correlation between the target industry theme and the target business object to be processed over Q time periods is determined based on the covariance, the industry standard deviation, and the business standard deviation.

[0208] The weights are determined based on the time periods corresponding to the relevance of each unit.

[0209] The Q-unit correlations are weighted and superimposed to form the correlation between the target industry theme and the target business object to be processed.

[0210] In one embodiment, the processor 1004 further performs the following steps:

[0211] Obtain multiple unit business trends of the target business object to be processed, and each unit business trend has a corresponding timestamp;

[0212] Based on the timestamp corresponding to each unit business trend, multiple unit business trends are divided into Q business trends. The timestamps of the unit business trends in the same business trend are in the same time period, and the unit business trends in the same business trend are arranged in ascending order of timestamp.

[0213] Obtain multiple unit industry trends for the target industry theme, with each unit industry trend corresponding to a timestamp;

[0214] Based on the timestamp corresponding to each unit industry trend, multiple unit industry trends are divided into Q industry trends. The timestamps of the unit industry trends within the same industry trend are in the same time period, and the unit industry trends within the same industry trend are arranged in ascending order of timestamp.

[0215] In one embodiment, when the processor 1004 determines the target business object for responding to the recommendation request from N target business objects based on N relevances, it specifically performs the following steps:

[0216] Sort the N target business objects in descending order of relevance, and select the first P target business objects as the target business objects to respond to the recommendation request, where P is a positive integer; or...

[0217] Target business objects with a relevance greater than a relevance threshold are selected as the target business objects for responding to the recommendation request; or...

[0218] Based on the business trend of each target business object, the business revenue of each target business object is determined. The N relevances and N business revenues are superimposed to obtain the recommendation coefficient of each target business object. The target business objects with recommendation coefficients greater than the recommendation threshold are used as the target business objects to respond to the recommendation request.

[0219] In one embodiment, the processor 1004 further performs the following steps:

[0220] The application page displays K industry topics. When a target industry topic is selected, a recommendation request is generated for that target industry topic; or...

[0221] When the popularity of a target industry topic exceeds a popularity threshold, a recommendation request is generated for the target industry topic.

[0222] When the processor 1004 executes the output of the target business object in response to the recommendation request, it specifically performs the following steps:

[0223] The application page displays the target business object used to respond to the recommendation request.

[0224] In one embodiment, the target business object is the target advertising data, the transaction object contained in the target business object is the promotional object contained in the target advertising data, and the transaction data volume of the promotional object is the promotional proportion of the promotional object in the advertising data.

[0225] When the processor 1004 determines the target business object for responding to the recommendation request from N target business objects based on N correlations, it specifically performs the following steps:

[0226] Obtain the conversion rate for each target ad data point, and then sum the N conversion rates and N relevances to obtain the recommendation coefficient for each target ad data point.

[0227] The target advertising data with a recommendation coefficient greater than the recommendation threshold will be used as the target advertising data to respond to the recommendation request.

[0228] In one embodiment, the target business object is the target text, and the transaction object included in the target business object is the keyword group of the target text. The transaction data volume of the keyword group is the frequency of the keyword group in the text.

[0229] When the processor 1004 determines the target business object for responding to the recommendation request from N target business objects based on N correlations, it specifically performs the following steps:

[0230] Obtain the attribute information for each target text, including one or more of the following: number of reposts, number of favorites, and number of comments;

[0231] The attribute information of each target text is combined with the relevance of each target text to form the input features of each target text;

[0232] The classification model is invoked to identify the N input features respectively, and the recommendation coefficient of each target text is obtained. The target text with the recommendation coefficient greater than the recommendation threshold is used as the target text to respond to the recommendation request.

[0233] In one embodiment, computer device 1000 is a first node, and the recommendation request is generated by a second node and sent to the first node;

[0234] When the processor 1004 retrieves N target business objects belonging to the target industry theme from M business objects, it specifically performs the following steps:

[0235] Obtain the block height corresponding to the target industry theme, extract the target block corresponding to the block height from the blockchain, and read N target business objects belonging to the target industry theme from the target block. The blockchain stores M business objects.

[0236] When the processor 1004 executes the output of the target business object in response to the recommendation request, it specifically performs the following steps:

[0237] The target business object used to respond to the recommendation request is sent to the second node.

[0238] It should be understood that the computer device 1000 described in the embodiments of this application can execute the foregoing text. Figures 3-8The description of the data processing method in the corresponding embodiments can also be performed as described above. Figure 9 The description of the data processing device 1 in the corresponding embodiments will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated here.

[0239] Furthermore, it should be noted that this application embodiment also provides a computer storage medium, which stores a computer program executed by the aforementioned data processing device 1. The computer program includes program instructions, and when the processor executes the program instructions, it can execute the aforementioned... Figures 3-8 The description of the data processing method in the corresponding embodiments is already provided and will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated. For technical details not disclosed in the embodiments of the computer storage medium involved in this application, please refer to the description of the method embodiments of this application. As an example, program instructions can be deployed and executed on a single computer device, or on multiple computer devices located in one location, or on multiple computer devices distributed across multiple locations and interconnected via a communication network. Multiple computer devices distributed across multiple locations and interconnected via a communication network can be combined to form a blockchain network.

[0240] According to one aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the aforementioned... Figures 3 to 8 The methods described in the corresponding embodiments are therefore not repeated here.

[0241] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0242] The above-disclosed embodiments are merely preferred embodiments of this application and should not be construed as limiting the scope of this application. Therefore, any equivalent variations made in accordance with the claims of this application shall still fall within the scope of this application.

Claims

1. A data processing method, characterized in that, include: When the popularity of a target industry topic is detected to be greater than a preset popularity threshold, a recommendation request for the target industry topic is generated, and N target business objects belonging to the target industry topic are extracted from the object industry pool. The object industry pool includes M business objects and the industry topic to which each business object belongs. Each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects. Each transaction object corresponds to an industry topic. The N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry topic corresponding to the transaction objects. N≤M, and N is a positive integer. The correlation between the target industry theme and each target business object is obtained. The target business object to be processed is one of N target business objects. The correlation between the target industry theme and the target business object to be processed is determined based on the covariance between the industry trend of the target industry theme and the business trend of the target business object to be processed, the industry standard deviation of the industry trend of the target industry theme, and the business standard deviation of the business trend of the target business object to be processed. Based on N correlations, determine the target business object for responding to the recommendation request from N target business objects, and output the target business object for responding to the recommendation request. Wherein, the number of industry trends for the target industry theme and the number of business trends for the target business object to be processed are both Q. The Q industry trends are obtained by dividing multiple unit industry trends of the target industry theme, where the timestamps of unit industry trends within the same industry trend are in the same time period, and the unit industry trends within the same industry trend are arranged in ascending order of timestamps. Similarly, the Q business trends are obtained by dividing multiple unit business trends of the target business object to be processed, where the timestamps of unit business trends within the same business trend are in the same time period, and the unit business trends within the same business trend are arranged in ascending order of timestamps. There is a one-to-one correspondence between the Q time periods of the Q industry trends and the Q time periods of the Q business trends; Q is an integer greater than 1; the correlation between the target industry theme and the target business object to be processed is determined based on the covariance, the industry standard deviation, and the business standard deviation, including: determining the unit correlation between the target industry theme and the target business object to be processed in the Q time periods based on the covariance, the industry standard deviation, and the business standard deviation; determining the weight according to the time period corresponding to each unit correlation; and superimposing the Q unit correlations using the weights of the Q unit correlations to obtain the correlation between the target industry theme and the target business object to be processed.

2. The method according to claim 1, characterized in that, M business objects contain transaction objects corresponding to K industry themes, and the target industry theme is one of the K industry themes; The method further includes: Get M business objects; Based on the transaction data volume of each business object and the industry theme corresponding to the transaction object, determine the industry theme to which each business object belongs, and associate and store each business object and the industry theme to which each business object belongs in the object industry pool. The step of obtaining N target business objects belonging to the target industry theme from M business objects includes: Extract N target business objects belonging to the target industry theme from the object industry pool.

3. The method according to claim 2, characterized in that, The pending business object is one of M business objects. The methods for determining the industry theme to which the pending business object belongs include: The transaction objects contained in the business objects to be processed are divided into one or more transaction sets, and the transaction objects in each transaction set belong to the same industry theme; The total transaction volume of each transaction set is calculated. The total transaction volume of the target transaction set is obtained by summing up the transaction data of the transaction objects in the target transaction set. The target transaction set is one of one or more transaction sets. The industry theme corresponding to the transaction set whose total transaction volume exceeds the transaction volume threshold is taken as the industry theme to which the business object to be processed belongs.

4. The method according to claim 1, characterized in that, The step of determining the target business object for responding to the recommendation request from N target business objects based on N relevances includes: Sort the N target business objects in descending order of relevance, and select the first P target business objects as the target business objects to respond to the recommendation request, where P is a positive integer; or... Target business objects with a relevance greater than a relevance threshold are selected as the target business objects for responding to the recommendation request; or... Based on the business trend of each target business object, the business revenue of each target business object is determined. The N relevances and N business revenues are superimposed to obtain the recommendation coefficient of each target business object. The target business objects with recommendation coefficients greater than the recommendation threshold are used as the target business objects to respond to the recommendation request.

5. The method according to claim 4, characterized in that, The method further includes: The application page displays K industry topics. When a target industry topic is selected, a recommendation request is generated for that target industry topic; or... When the popularity of a target industry topic exceeds a popularity threshold, a recommendation request is generated for the target industry topic. The output is used to respond to the target business object of the recommendation request, including: The application page displays the target business object used to respond to the recommendation request.

6. The method according to claim 1, characterized in that, The target business object is the target advertising data. The transaction object contained in the target business object is the advertising object contained in the target advertising data. The transaction data volume of the advertising object is the advertising proportion of the advertising object in the advertising data. The step of determining the target business object for responding to the recommendation request from N target business objects based on N relevances includes: Obtain the conversion rate for each target ad data point, and then sum the N conversion rates and N relevances to obtain the recommendation coefficient for each target ad data point. The target advertising data with a recommendation coefficient greater than the recommendation threshold will be used as the target advertising data to respond to the recommendation request.

7. The method according to claim 1, characterized in that, The target business object is the target text, and the transaction object included in the target business object is the keyword group of the target text. The transaction data volume of the keyword group is the frequency of the keyword group in the text. The step of determining the target business object for responding to the recommendation request from N target business objects based on N relevances includes: Obtain the attribute information for each target text, including one or more of the following: number of reposts, number of favorites, and number of comments; The attribute information of each target text is combined with the relevance of each target text to form the input features of each target text; The classification model is invoked to identify the N input features respectively, and the recommendation coefficient of each target text is obtained. The target text with the recommendation coefficient greater than the recommendation threshold is used as the target text to respond to the recommendation request.

8. The method according to claim 1, characterized in that, The method is executed by the first node, and the recommendation request is generated by the second node and sent to the first node; The step of obtaining N target business objects belonging to the target industry theme from M business objects includes: Obtain the block height corresponding to the target industry theme, extract the target block corresponding to the block height from the blockchain, and read N target business objects belonging to the target industry theme from the target block. The blockchain stores M business objects. The output is used to respond to the target business object of the recommendation request, including: The target business object used to respond to the recommendation request is sent to the second node.

9. A data processing apparatus, characterized in that, include: The first acquisition module is used to generate a recommendation request for the target industry theme when the detected popularity of the target industry theme is greater than a preset popularity threshold, and to extract N target business objects belonging to the target industry theme from the object industry pool; wherein, the object industry pool includes M business objects and the industry theme to which each business object belongs, each of the M business objects contains one or more transaction objects and the transaction data volume of the transaction objects, each transaction object corresponds to an industry theme, and the N target business objects are selected from the M business objects based on the transaction data volume of the transaction objects contained in each business object and the industry theme corresponding to the transaction objects, where N≤M, and N is a positive integer; The second acquisition module is used to acquire the correlation between the target industry theme and each target business object. The target business object to be processed is one of N target business objects. The correlation between the target industry theme and the target business object to be processed is determined based on the covariance between the industry trend of the target industry theme and the business trend of the target business object to be processed, the industry standard deviation of the industry trend of the target industry theme, and the business standard deviation of the business trend of the target business object to be processed. The selection module is used to determine the target business object for responding to the recommendation request from N target business objects based on N relevances, and output the target business object for responding to the recommendation request. Wherein, the number of industry trends of the target industry theme and the number of business trends of the target business object to be processed are both Q. The Q industry trends are obtained by dividing multiple unit industry trends of the target industry theme, where the timestamps of the unit industry trends in the same industry trend are in the same time period, and the unit industry trends in the same industry trend are arranged in ascending order of timestamps; the Q business trends are obtained by dividing multiple unit business trends of the target business object to be processed, where the timestamps of the unit business trends in the same business trend are in the same time period, and the unit business trends in the same business trend are arranged in ascending order of timestamps; the Q time periods of the Q industry trends and the Q time periods of the Q business trends correspond one-to-one; Q is an integer greater than 1; the second acquisition module is specifically used to: determine the unit correlation between the target industry theme and the target business object to be processed in the Q time periods based on the covariance, the industry standard deviation, and the business standard deviation; determine the weight according to the time period corresponding to each unit correlation; and use the weights of the Q unit correlations to superimpose the Q unit correlations into the correlation between the target industry theme and the target business object to be processed.

10. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method according to any one of claims 1-8.

11. A computer storage medium, characterized in that, The computer storage medium stores a computer program, the computer program including program instructions, which, when executed by a processor, cause a computer device having the processor to perform the steps of the method according to any one of claims 1-8.

12. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method according to any one of claims 1-8.

Images