Information batch pushing method and platform and computer storage medium

Abstract

The invention relates to the technical field of information processing, in particular to an information batch pushing method and platform and a computer storage medium, and the method comprises the steps: firstly responding to a user request, obtaining a candidate data set, and then executing primary filtering based on a uniform identity label and bottom filtering based on a user name on the candidate data set; the method comprises the steps of generating a candidate list after duplicate removal, then verifying the current interaction state of each candidate in the candidate list to reject candidates in an interacted state to obtain a target push list, and finally pushing corresponding invitation information to an object recorded in the target push list. Compared with the prior art, the method has the advantages that double deduplication strategies of'primary filtering 'and'bottom filtering' are integrated, the uniqueness of the pushed object in a high-concurrency environment can be ensured, and therefore the problem of repeated pushing is avoided.

Description

Technical Field

[0001] This invention relates to the field of information processing technology. More specifically, this invention relates to a method, platform, and computer storage medium for batch information delivery. Background Technology

[0002] In the fields of cross-border e-commerce and influencer marketing, merchants typically need to send bulk collaboration invitations to a large number of influencers. Existing bulk push technologies usually only offer simple list import and bulk sending functions, and still have the following shortcomings: First, the lack of global state constraints means the system cannot identify the status of the same influencer in different campaigns. For example, influencer A may have already responded to a collaboration request in "Campaign 1," but in this concurrent interaction scenario, the merchant may not be able to know in time and send another invitation in "Campaign 2," which could cause information overload and even lead to influencer churn.

[0003] Secondly, there are conflicts in the identification of multiple data sources. Influencer data may come from different channels or media platforms. When information is returned, some returned information has a unique ID, but others only have a username. Using traditional simple deduplication logic, that is, matching only the ID or only the username, is extremely easy to lead to information omissions and duplicate associations.

[0004] Therefore, existing batch information push technologies face the technical challenge of repeated pushes in concurrent control scenarios. Summary of the Invention

[0005] To address the aforementioned technical problem of repeated push notifications, this invention discloses a method, platform, and computer storage medium for batch information push notifications.

[0006] In a first aspect, the present invention discloses a method for batch information push, comprising: In response to a user request, retrieve the candidate dataset; the candidate data in the candidate dataset comes from different data sources; Perform primary filtering based on unified identity and fallback filtering based on user name on the candidate dataset to generate a deduplicated candidate list; Verify the current interaction status of each candidate in the candidate list to remove candidates who are already in an interactive state, and obtain the target push list; Send the corresponding invitation information to the objects recorded in the target push list.

[0007] Preferably, the candidate dataset is subjected to primary filtering based on a unified identity and fallback filtering based on the user name, including: Extract the uniform identity identifier from the candidate dataset; Search the existing data for the existence of the same identifier. If the same identifier is found, mark the corresponding candidate as a duplicate and remove the corresponding candidate data from the candidate dataset. Extract the usernames of the remaining candidates after removal; Search the existing data to see if there are any duplicate usernames. If so, mark the corresponding candidate as a duplicate and remove the corresponding candidate data from the candidate dataset. The candidate dataset, after two rounds of elimination, is reconstructed to generate a candidate list.

[0008] Preferably, in response to a user request, the candidate dataset is obtained, including: The system uses a rule engine or artificial intelligence model to analyze user requests and obtain candidate screening criteria and user permissions to access third-party data. Based on the third-party data access permissions, call the corresponding third-party API to retrieve candidates according to the filtering conditions, and summarize the retrieval results into a candidate dataset.

[0009] Preferably, after generating the candidate dataset, the method of the present invention further includes: Iterate through the candidate dataset to check if the contact information of each candidate is complete. If not, an asynchronous data acquisition command is generated, which triggers a contact information completion operation via a message queue; For entities that already have valid contact information or have completed the necessary documentation, batch-wise data entry will be performed.

[0010] Preferably, the completion operation is configured as follows: Obtain the corresponding contact information from multiple authorized third-party platforms and fill in the corresponding missing information fields.

[0011] Preferably, after pushing invitation information to objects recorded in the target push list, the method of the present invention further includes: The lifecycle state of the initialized object is set to the invited state; Monitor the interaction behavior of objects in real time to update the current interaction state of the objects.

[0012] Preferably, the current interaction status includes at least no invitation, no reply, invitation, and reply.

[0013] Preferably, invitation messages are sent via private message and email.

[0014] Secondly, the present invention also discloses an information batch push platform for the information batch push method described in the first aspect. The platform of the present invention includes: The user interaction layer is used to receive user requests; The control layer is used to respond to user requests and issue drive commands; The service layer is used to receive driver instructions and execute the following steps: In response to a user request, retrieve the candidate dataset; the candidate data in the candidate dataset comes from different data sources; Perform primary filtering based on unified identity and fallback filtering based on user name on the candidate dataset to generate a deduplicated candidate list; Verify the current interaction status of each candidate in the candidate list to remove candidates who are already in an interactive state, and obtain the target push list; Send the corresponding invitation information to the objects recorded in the target push list.

[0015] Thirdly, the present invention also discloses a computer storage medium, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the information batch push method described in the first aspect is implemented.

[0016] The beneficial effects of this invention are as follows: (1) Compared with the prior art, the method of the present invention incorporates a dual deduplication strategy of "primary filtering" and "backup filtering", which can ensure the uniqueness of the push object in a high-concurrency environment, thereby avoiding the problem of duplicate push.

[0017] (2) Compared with the prior art, the method of the present invention can summarize the object association information of multi-source data to obtain a more reliable and accurate target push list, thereby further reducing the possibility of duplicate push.

[0018] (3) Compared with the prior art, the method of the present invention can also link multiple third-party platforms to perform asynchronous data completion, ensuring the integrity of the contact information of the connected objects, thereby avoiding the possibility of invalid push. Attached Figure Description

[0019] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein: Figure 1 This is a flowchart of the information batch push method in Embodiment 1 of the present invention; Figure 2 This is an architecture diagram of the information batch push system in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the structure of the computer storage medium in Embodiment 3 of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0022] Example 1 like Figure 1 As shown, this embodiment discloses a method for batch information push, including: S10: Retrieve candidate dataset in response to user request.

[0023] It should be noted that the same candidate data in the candidate dataset may come from different data sources. These different data sources could be TikTok, X, INS, YouTube, and Instagram, among others.

[0024] Further, step S10 above includes: S11: Use a rule engine or artificial intelligence model to parse user requests and obtain candidate screening conditions and user's third-party data access permissions.

[0025] It should be noted that the above filtering criteria may include the range of the number of followers, the range of interaction rate, region, category, and whether or not an email address is available.

[0026] In one embodiment, a user can input filtering criteria into the rules engine, and then use the rules engine to identify and call APIs with third-party permissions to retrieve candidate datasets from various media platforms to obtain the candidate dataset.

[0027] In another embodiment, the user can input a natural language statement into the artificial intelligence model (NLP model), such as "I want to send invitation information to major mainstream media platforms". After the natural language is parsed, specific filtering conditions can be obtained, and the user's third-party API authorization can be obtained based on the user's own tags or ID, and then step S12 can be executed.

[0028] S12: Based on the third-party data acquisition permissions, call the corresponding third-party API to retrieve candidates according to the filtering conditions, and summarize the retrieval results into a candidate dataset.

[0029] The aggregated candidate dataset can be temporarily stored in Redis using a matrix vector or a table. Each row of the candidate dataset records the candidate's username, unique identifier (ID), interaction status, active time, contact email, and contact phone number.

[0030] To address the situation where subsequent invitations cannot be successfully made due to the lack of a contact email or phone number, in this embodiment, after step S12, the method further includes: S100: Iterate through the candidate dataset to check if the contact information of each candidate is complete.

[0031] Specifically, iterate through each row of the candidate dataset to check if the contact email and phone number are empty. If they are, it means that the contact information is not complete, and proceed to step S101. If not, proceed to step S102.

[0032] S101: If not, generate an asynchronous data acquisition command and trigger a contact information completion operation through the message queue.

[0033] Furthermore, the above completion operation can be implemented by a pre-configured script. When step S100 triggers the "No" condition, the script automatically performs the completion operation for the specific row data. This script is pre-configured as follows: The system retrieves corresponding contact information from multiple authorized third-party platforms and fills in the missing information fields. If none of the authorized third-party platforms can retrieve the corresponding contact information, an error feedback is triggered to notify the user to manually confirm or interactively complete the input.

[0034] S102: For objects that already have valid contact information or have completed the necessary information, perform batch entry operations.

[0035] It should be explained that in a high-concurrency control system, the process of establishing invitations and interactions between users and candidates is continuous and synchronous. In this scenario, it is necessary to avoid excessive pressure on the database's input / output. Therefore, for the insertion of large amounts of object data, this embodiment adopts a batch insertion method.

[0036] S20: Perform primary filtering based on unified identity and fallback filtering based on username on the candidate dataset to generate a deduplicated candidate list.

[0037] Specifically, step S20 above includes: S21: Extract the uniform identity identifier from the candidate dataset.

[0038] In this embodiment, the aforementioned unified identity identifier mainly refers to the candidate's account ID.

[0039] S22: Search the existing data for the existence of the same identifier. If so, mark the corresponding candidate as a duplicate and remove the corresponding candidate data from the candidate dataset.

[0040] It should be explained that the execution of step S22 above is to more quickly filter out influencers who have already successfully established connections, and to ensure that when invitation information is pushed, there is less chance of repeated pushes to influencers who have already established connections, thereby reducing information harassment to influencers and thus initially solving the technical problem of repeated pushes in the existing technology.

[0041] Considering that account ID formats vary in multi-platform data access scenarios, the same influencer may have multiple different account IDs on different media platforms. Therefore, in addition to step S22 above, step S23 also needs to be executed.

[0042] S23: Extract the usernames of the remaining candidates after removal.

[0043] Preferably, the aforementioned username and phone number are bound together; therefore, while retrieving the username, the phone number to which it is bound is also retrieved.

[0044] S24: Search the existing data for the existence of the same username. If so, mark the corresponding candidate as a duplicate and remove the corresponding candidate data from the candidate dataset.

[0045] It should be noted that although identical usernames are not allowed on the same media platform, in cross-platform multi-source data access scenarios, it is quite common for the same username to appear on multiple platforms. Therefore, in step S24, the phone number bound to the username can also be verified. If the same username and bound number exist in the existing data, it means that the influencer corresponding to the username has already established a connection, and there is no need to push invitation information again.

[0046] S25: Reconstruct the information of the candidate dataset after two rounds of elimination to generate a candidate list.

[0047] Through steps S21-S24 above, the method of this embodiment introduces a dual deduplication strategy of "primary filtering" and "fallback filtering". As the most core strategy in the method of this embodiment, it can ensure the uniqueness of the push object in a high-concurrency environment, thereby further avoiding the problem of duplicate push.

[0048] The data pruning process involves database operations, such as table operations in MySQL. Once the pruning is complete, empty rows are deleted, and the data is reorganized, sorted, and aggregated into a candidate list. To improve the success rate of invitations, weights can be assigned to dimensions such as follower count range, interaction rate range, region, and category when sorting the list. These weights are used to assign values ​​to quantifiable metrics, identifying higher-value candidates and prioritizing them to facilitate message delivery.

[0049] S30: Verify the current interaction state of each candidate in the candidate list to remove candidates who are already in an interactive state, and obtain the target push list.

[0050] In this embodiment, the current interaction status includes at least no invitation, no reply, invitation, and reply.

[0051] It's important to further explain that in some special scenarios, even after passing the dual deduplication of the aforementioned "primary filtering" and "fallback filtering," omissions may still occur. For example, a candidate might have registered multiple accounts on multiple platforms using multiple phone numbers. In such cases, account association needs to be performed beforehand. If step S20 fails, after generating the candidate list, the push accounts in the historical database are retrieved to verify whether there are associated accounts in the candidate list. If so, the corresponding candidate data in the target push list is removed.

[0052] S40: Push the corresponding invitation information to the objects recorded in the target push list.

[0053] The invitation information can be sent via private message, SMS, or email. For private messages, this is an online interactive invitation from the corresponding media platform. After identifying the recipients, a script uses the official operating account of the corresponding media platform to generate a private message template and send it to each recipient. Similarly, for emails, after identifying the recipients, a script uses an official email template to forward emails in bulk to the recipients in the target list via the official operating email address.

[0054] Furthermore, after pushing the invitation information to the objects recorded in the target push list, the method of this embodiment also includes: S500: The lifecycle state of the initialized object is set to the invited state.

[0055] S501: Monitors the interaction behavior feedback of objects in real time to update the current interaction state of the objects.

[0056] Specifically, a preset state machine is used to monitor the interactive behavior feedback of the object in real time and send information feedback to the corresponding push platform to update the current interactive state of the object.

[0057] In summary, compared with the prior art, the method of this embodiment has at least the following advantages: Firstly, the method in this embodiment has an intelligent deduplication mechanism, which can ensure the uniqueness of the connection or invitation with the influencer, greatly reduce the possibility of repeated pushes in concurrent scenarios, and reduce information harassment to the influencer.

[0058] Secondly, based on the dual filtering, the method in this embodiment supports cross-Campaign (promotion plan) constraint filtering through step S30, which can avoid duplicate invitations from a global perspective and further improve the connection building experience for influencers.

[0059] Third, the method in this embodiment supports the integration of multiple data sources, which can integrate the influencer data resources of multiple third-party media platforms and dynamically update the influencer data.

[0060] Fourth, the method in this embodiment supports asynchronous data completion, which can complete the contact information of experts through asynchronous MQ tasks, thereby reducing the blocking of the main process, i.e., the push task.

[0061] Fifth, the method in this embodiment supports batch insertion of large amounts of data, which can reduce the pressure on the database in high-concurrency scenarios.

[0062] Sixth, the method in this embodiment introduces a state machine management mechanism, which can fully monitor the state changes of each expert, thereby accurately tracking the collaboration progress of each expert.

[0063] Example 2 like Figure 2 As shown, this embodiment discloses an information batch push platform for use with the information batch push method described in the first aspect. The aforementioned information batch push platform includes a user interaction layer, a control layer, a service layer, and a mapping layer.

[0064] in, The user interaction layer is used to receive user requests.

[0065] The control layer is used to respond to user requests and issue drive commands.

[0066] The service layer is used to receive driver instructions and execute the following steps: In response to a user request, a candidate dataset is retrieved; the candidate data in the candidate dataset comes from different data sources.

[0067] Perform primary filtering based on unified identity and fallback filtering based on username on the candidate dataset to generate a deduplicated candidate list.

[0068] Verify the current interaction status of each candidate in the candidate list to remove candidates who are already in an interactive state, and obtain the target push list.

[0069] Send the corresponding invitation information to the objects recorded in the target push list.

[0070] It should be noted that the cached data generated during the execution of the aforementioned service layer is stored in a Redis cache to support high-concurrency scenarios and thus alleviate database pressure. The Redis cache is also configured with distributed locks to ensure that only one operation is possible at a time when multiple services or processes access shared resources simultaneously, thereby preventing data corruption or inconsistency.

[0071] To achieve a more efficient push platform, when the aforementioned invitation message push task is executed, the RocketMQ platform will selectively execute email or private message sending methods and dynamically update the status of the push recipient. The RocketMQ platform is an open-source distributed messaging and streaming data platform known for its low latency, high performance, and high reliability. It supports trillions of message backlogs and is suitable for asynchronous communication, decoupling, and traffic shaping scenarios in large-scale distributed systems.

[0072] In the task allocation and scheduling process at the service layer, the platform in this embodiment also incorporates XXL-JOB, which supports flexible distribution and execution of push tasks within the cluster to avoid single points of failure. Specifically, XXL-JOB acts as the task scheduling center, responsible for receiving push task configurations, allocating execution nodes, and monitoring the status of push tasks. It avoids service interruptions caused by expired logins from third-party authorized platforms by periodically updating system access tokens, making it a crucial task scheduling tool for maintaining the continuous automatic operation of the platform in this embodiment.

[0073] The mapping layer is used to interface with the service layer mentioned above and perform operations such as storing, deleting, or retrieving data from the database (MySQL) to meet the operational needs of the service layer.

[0074] Example 3 like Figure 3 As shown, this embodiment also discloses a computer storage medium, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the information batch push method described in the first aspect is implemented.

[0075] The aforementioned computer storage media also includes other components well known to those skilled in the art, such as communication interfaces, the settings and functions of which are known in the art and therefore will not be described in detail here.

[0076] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions that can be stored or otherwise maintained by such a computer-readable medium.

[0077] In the description of this specification, "multiple" means at least two, such as two, three or more, etc., unless otherwise expressly and specifically defined.

[0078] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A method for batch information push, characterized in that, include: In response to a user request, a candidate dataset is obtained; wherein the candidate data in the candidate dataset comes from different data sources; Perform primary filtering based on unified identity and fallback filtering based on user name on the candidate dataset to generate a deduplicated candidate list; Verify the current interaction state of each candidate in the candidate list to remove candidates who are already in an interactive state, and obtain the target push list; Send the corresponding invitation information to the objects recorded in the target push list.

2. The information batch push method according to claim 1, characterized in that, Perform primary filtering based on unified identity and fallback filtering based on username on the candidate dataset, including: Extract the unified identity identifier from the candidate dataset; Search the existing data for the existence of the same identifier. If the same identifier is found, mark the corresponding candidate as a duplicate and remove the corresponding candidate data from the candidate dataset. Extract the usernames of the remaining candidates after removal; Search the existing data to see if there are any duplicate usernames. If so, mark the corresponding candidate as a duplicate and remove the corresponding candidate data from the candidate dataset. The candidate dataset after two rounds of elimination is reconstructed to generate the candidate list.

3. The information batch push method according to claim 1, characterized in that, In response to a user request, retrieve the candidate dataset, including: The system uses a rule engine or artificial intelligence model to analyze user requests and obtain candidate screening criteria and user permissions to access third-party data. Based on the third-party data access permissions, the corresponding third-party API is called to retrieve candidates according to the filtering conditions, and the retrieval results are summarized into the candidate dataset.

4. The information batch push method according to claim 1, characterized in that, After generating the candidate dataset, the method further includes: Iterate through the candidate dataset to check if the contact information of each candidate is complete. If not, an asynchronous data acquisition command is generated, which triggers a contact information completion operation via a message queue; For those who already have valid contact information or have completed the necessary steps, perform batch-by-batch data entry.

5. The information batch push method according to claim 4, characterized in that, The completion operation is configured as follows: Obtain the corresponding contact information from multiple authorized third-party platforms and fill in the corresponding missing information fields.

6. The method for batch information push according to claim 1, characterized in that, After sending the invitation information to the objects recorded in the target push list, the method further includes: The lifecycle state of the initialized object is set to the invited state; Monitor the interaction behavior of objects in real time to update the current interaction state of the objects.

7. The method for batch information push according to claim 1, characterized in that, The current interaction status includes at least no invitation, no reply, invitation, and reply.

8. The method for batch information push according to claim 1, characterized in that, The invitation information can be sent via private message or email.

9. An information batch push platform, characterized in that, The platform for the information batch push method according to any one of claims 1-8 includes: The user interaction layer is used to receive user requests; The control layer is used to respond to the user request and issue drive commands; The service layer is used to receive the driver instructions and execute the following steps: In response to a user request, a candidate dataset is obtained; wherein the candidate data in the candidate dataset comes from different data sources; Perform primary filtering based on unified identity and fallback filtering based on user name on the candidate dataset to generate a deduplicated candidate list; Verify the current interaction state of each candidate in the candidate list to remove candidates who are already in an interactive state, and obtain the target push list; Send the corresponding invitation information to the objects recorded in the target push list.

10. A computer storage medium, characterized in that, It includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the information batch push method according to any one of claims 1-8 is implemented.

Images