Method, device, storage medium and equipment for processing financial product operation report

Abstract

The application discloses a kind of financial product operation report processing method, device, storage medium and equipment, it is related to financial technology technical field, can improve the processing efficiency and processing accuracy of financial product operation report. Including: configuration report receiving rule;Based on report receiving rule, automatically collect financial product operation report from multiple preset sources, and the collected financial product operation report is uniformly analyzed and structured processing, generates the structured product label containing product unique identification, and the structured product label is associated with the financial product operation report Storage;Determine the target user set with report viewing permission, and obtain the environment fingerprint information of each target user using terminal in target user set;Financial product operation report is disassembled into multiple logical subflows, and the personalized streaming strategy is generated for each target user based on environment fingerprint information, and the logical subflow is pushed to the using terminal of corresponding target user based on streaming strategy.

Description

Technical Field

[0001] This invention relates to the field of financial technology, and in particular to a method, apparatus, storage medium, and device for processing financial product operation reports. Background Technology

[0002] A financial product operation report is a document provided by the product manager to investors after they purchase a financial product from a brokerage firm, detailing the product's operation. Based on these reports, brokerage firms provide clients with services such as product operation information disclosure, answering portfolio inquiries, providing suitability adjustment suggestions, and offering support for protecting their rights. The brokerage firm needs to process these operation reports during the service process.

[0003] Currently, financial product operation reports are typically processed manually. However, this manual method is time-consuming and labor-intensive, and errors can occur due to varying skill levels among staff or oversight. Summary of the Invention

[0004] This invention provides a method, apparatus, storage medium, and device for processing financial product operation reports, which mainly improves the processing efficiency and accuracy of financial product operation reports.

[0005] According to a first aspect of the present invention, a method for processing financial product operation reports is provided, comprising: In response to the processing signal of the financial product operation report, a report receiving rule is configured, wherein the report receiving rule includes at least one of the following: report source address, report identification regular expression, and report synchronization frequency; Based on the report receiving rules, financial product operation reports are automatically collected from multiple preset sources. The collected financial product operation reports are uniformly parsed and structured to generate structured product tags containing unique product identifiers. The financial product operation reports are associated with and stored with the structured product tags. The preset sources include at least one of email, internal marketing system, and file upload interface. A set of target users with report viewing permissions is determined, and environmental fingerprint information of the terminal used by each target user in the set of target users is obtained, wherein the environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status; Based on the content of the financial product operation report stored in the association, the financial product operation report is decomposed into multiple logical sub-streams, and a personalized streaming transmission strategy is generated for each target user based on the environmental fingerprint information. The logical sub-streams are then pushed to the corresponding target user's terminal based on the streaming transmission strategy. The logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream. The streaming transmission strategy defines at least one of the following for each logical sub-stream: transmission order, compression ratio, and start / stop control rules.

[0006] Optionally, based on the content of the financial product operation report stored in the associated database, the financial product operation report is decomposed into multiple logical sub-streams, including: Extract the key numerical indicators and text summaries from the financial product operation report, and encapsulate the key numerical indicators and text summaries into a core data stream; The system identifies the charts, curves, and image elements in the financial product operation report, converts the charts, curves, and image elements into vector data, and encapsulates the vector data into a visual chart data stream. The entire financial product operation report is packaged into a full-document archive data stream.

[0007] Optionally, a personalized streaming strategy is generated for each target user based on the environmental fingerprint information, including: If it is detected that the user terminal is in a weak network environment and the user interface focus is on the report details page, the streaming strategy is set to transmit the core data stream at full speed only, the visualization chart data stream is downgraded to low-resolution thumbnail transmission, and the transmission of the full document file data stream is suspended. If it is detected that the terminal is in a high-bandwidth environment and the user interface focus is on the message list, the streaming strategy is set to silently preload the core data stream, visual icon stream, and full document file data stream to the local cache in the background; If the battery level of the terminal is detected to be below a preset threshold, the streaming policy is set to limit the frame rate or resolution of the visualization chart data stream.

[0008] Optionally, determine the target set of users with report viewing permissions, including: The system identifies multi-dimensional correlation information between multiple candidate users and the financial product operation report, user characteristic information of the candidate users, and report characteristic information of the financial product operation report. The multi-dimensional correlation information includes the candidate users' holding characteristics, behavioral characteristics, and implicit correlations with the financial products in the financial product operation report. Based on the multi-dimensional association information, the user feature information, and the report feature information, a permission score is generated for each candidate user, and a set of target users with report viewing permissions is matched among each candidate user based on the permission score.

[0009] Optionally, the method further includes: The processing of financial product operation reports generates processing logs with transaction tracking identifiers. The processing process includes at least one of the following: configuration process for report receiving rules, collection process, parsing process, structured processing process, storage process, logical sub-stream decomposition process, and logical sub-stream push process. Based on the processing logs, abnormal events in the financial product operation report processing process are identified, and root cause analysis is performed on the abnormal events. Based on the root cause analysis results, the abnormal type of the abnormal event is determined, an abnormal notification strategy is determined based on the abnormal type, and the abnormal event is pushed based on the abnormal notification strategy.

[0010] Optionally, after pushing the logical sub-stream to the terminal of the corresponding target user based on the streaming strategy, the method further includes: A behavior probe is deployed on the target user's terminal to capture interaction events in real time when the target user reads the financial product operation report. The interaction events include at least one of the following: page scrolling speed, page scrolling acceleration, chart zoom level change, text selection, text copying, cursor or touch hover trajectory, and cursor or touch dwell time. The target user's current reading behavior pattern is identified based on the interaction event, wherein the current reading behavior pattern includes one of the following: in-depth analysis mode, skimming and filtering mode, transaction intention mode, and confusion and hesitation mode. A dynamic content heat map is generated based on the current reading behavior pattern of each target user. High-attention content segments in the financial product operation report are determined based on the dynamic content heat map. The streaming transmission strategy for subsequent users reading the financial product operation report is adjusted based on the high-attention content segments. The adjustment of the streaming transmission strategy includes at least one of the following: automatically rearranging the chapter order, preloading high-attention data blocks, and automatically associating interpretation auxiliary information with high-difficulty content segments.

[0011] Optionally, the structured product tags include at least one of the following: product code, product name, report time, net asset value trend, sales rules, and product interpretation. Before storing the financial product operation report in association with the structured product tag, the method further includes: Using the product code as an index key, retrieve the standard benchmark data corresponding to the financial product operation report in a preset standardized financial database. The standard benchmark data includes the standard product name, standard report time, standard net asset value trend, standard sales rules, and standard product interpretation. The structured product label is verified for accuracy using the standard benchmark data. A data correction strategy is generated based on the accuracy verification results, and the structured product label is corrected using the data correction strategy.

[0012] According to a second aspect of the present invention, a processing apparatus for financial product operation reports is provided, comprising: A configuration unit is configured to configure report receiving rules in response to a processing signal for a financial product operation report, wherein the report receiving rules include at least one of a report source address, a report identification regular expression, and a report synchronization frequency; The storage unit is used to automatically collect financial product operation reports from multiple preset sources based on the report receiving rules, and to uniformly parse and structure the collected financial product operation reports to generate a structured product tag containing a unique product identifier. The financial product operation report is then associated with and stored with the structured product tag. The preset sources include at least one of email, internal marketing system, and file upload interface. The acquisition unit is used to determine a set of target users with report viewing permissions and acquire environmental fingerprint information of the terminal used by each target user in the set of target users, wherein the environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status. The information push unit is used to decompose the financial product operation report into multiple logical sub-streams based on the content of the associated stored financial product operation report, and generate a personalized streaming transmission strategy for each target user based on the environmental fingerprint information. Based on the streaming transmission strategy, the logical sub-streams are pushed to the corresponding target user's terminal. The logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream. The streaming transmission strategy defines at least one of the following: transmission order, compression ratio, and start / stop control rules for each logical sub-stream.

[0013] According to a third aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the above-described method for processing financial product operation reports.

[0014] According to a fourth aspect of the present invention, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-mentioned method for processing financial product operation reports.

[0015] The present invention provides a method, apparatus, storage medium, and device for processing financial product operation reports. Compared with the current method of manually processing financial product operation reports, the present invention, by configuring report receiving rules, is compatible with multiple operation report sources simultaneously. It can automatically monitor multiple channels without requiring manual login to the system to collect reports one by one, completely solving the problems of low efficiency and easy omissions in manual collection. Furthermore, it uses regular expressions to accurately match key information in file names or email titles, effectively filtering spam and irrelevant files, ensuring data purity from the source. Different fund managers send reports at different frequencies, and the configurable synchronization frequency allows the system to flexibly adapt to various business scenarios, without excessively consuming resources or missing urgent reports. By aggregating reports scattered from different sources to a unified platform, data fragmentation is eliminated. This system facilitates centralized management and uses a unique product identifier for associated storage, effectively preventing duplicate data from multiple uploads or receptions of the same report, ensuring database cleanliness and uniqueness. By defining the target user set with report viewing permissions, sensitive financial reports are ensured to be distributed only to compliant holders or authorized personnel, reducing the risk of information leakage. By dynamically adjusting the transmission strategy according to the real-time environment, loading failures caused by network fluctuations or device incompatibility are significantly reduced. For high-concurrency scenarios, this invention splits the report into streams, allowing caching and acceleration only for frequently accessed "core data," reducing peak bandwidth pressure on large files. Through the streaming transmission strategy, authentication and distribution can be achieved immediately after report parsing, significantly improving timeliness compared to the traditional manual distribution model. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 A flowchart illustrating a method for processing financial product operation reports according to an embodiment of the present invention is shown. Figure 2 A flowchart illustrating another method for processing financial product operation reports provided by an embodiment of the present invention is shown; Figure 3 This diagram illustrates the structure of a financial product operation report processing device according to an embodiment of the present invention. Figure 4A schematic diagram of the structure of another financial product operation report processing device provided in an embodiment of the present invention is shown; Figure 5 A schematic diagram of the physical structure of a computer device provided in an embodiment of the present invention is shown. Detailed Implementation

[0017] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.

[0018] Currently, processing financial product operation reports manually is time-consuming and labor-intensive, and errors may occur due to varying skill levels or negligence among staff.

[0019] To address the aforementioned problems, embodiments of the present invention provide a method for processing financial product operation reports, such as... Figure 1 As shown, the method includes: 101. In response to the processing signal of the financial product operation report, configure the report receiving rules, wherein the report receiving rules include at least one of the following: report source address, report identification regular expression, and report synchronization frequency.

[0020] In this embodiment of the invention, the processing method for financial product operation reports is implemented through a financial product operation report processing system. Staff members click the "Add Report Source" button on the "Report Source Configuration" interface of the processing system's management backend. The system then generates a processing signal for a financial product operation report, initializes the configuration process, and the staff member enters a specific report source address on the system's interactive interface, such as the official disclosure email address of a fund company. The system configures report recognition regular expressions; for example, the email title format of the fund company is fixed as: "[Operation Report] {Product Name}_{Report Date}_Month / Quarterly". The staff member enters the regular expression in the "Title Matching Rules" field: ^[Operation Report]._(\d{4}-\d{2}-\d{2})_(Monthly|Quarterly|Annual).$. The system configures the report synchronization frequency; for example, for managers who send daily net asset value reports, the user sets the synchronization frequency to "every 15 minutes" or "real-time monitoring". The system polls the email address every 15 minutes or maintains a long connection to push new email notifications instantly. Then, according to the set synchronization frequency, a scheduled task or listening thread is started to connect to the specified report source address, obtain the new email list, and apply report recognition regular expressions to match each email. For emails that match successfully, the system automatically extracts the attachments and proceeds to the subsequent parsing process; emails that fail to match are logged and skipped. In this way, the automatic collection of financial operation reports can be achieved. This embodiment of the invention can simultaneously support multiple operation report sources by configuring report receiving rules, and can automatically monitor multiple channels without the need for manual login to the system to capture each report. This completely solves the problems of low efficiency and easy omissions in manual collection. Moreover, by using regular expressions to accurately match key information in file names or email titles, it effectively filters spam and irrelevant files, ensuring the purity of data from the source. Different fund managers send reports at different frequencies, and the configurable synchronization frequency allows the system to flexibly adapt to various business scenarios, without excessively consuming resources or missing urgent reports.

[0021] 102. Based on the report receiving rules, automatically collect financial product operation reports from multiple preset sources, and uniformly parse and structure the collected financial product operation reports to generate structured product tags containing unique product identifiers. Store the financial product operation reports and structured product tags together. The preset sources include at least one of email, internal marketing system, and file upload interface.

[0022] Among them, the unique identifier of a product can be the product code; the structured product tags include, but are not limited to, product elements such as the product name, reporting time, net asset value trend, sales rules, and product interpretation in the financial product operation report.

[0023] In this embodiment of the invention, during the collection of financial product operation reports from email, the system first connects to the configured mail server via a protocol, scans email titles and bodies according to a preset "report recognition regular expression," filters target emails, automatically downloads attachments from matching emails, and extracts sender, sending time, and email subject as metadata. The downloaded attachments are then scanned for viruses, and if successful, stored in a temporary buffer along with the metadata. During the collection of financial product operation reports from the internal marketing system, the system first calls the standard API interface of the internal marketing system, requesting a new report list with an authentication token, and receives the file index information returned by the interface (including product code, report type, and upload time). The binary file stream is downloaded according to the index, and the data source is marked as "internal synchronization," thus obtaining the financial product operation report. During the collection of financial product operation reports from the file upload interface, the system opens the file upload interface to receive report files manually dragged and dropped by operations personnel.

[0024] Furthermore, for the large number of financial product operation reports collected, the report file type is checked. If it is in Word, Excel, or image format, a conversion service is called to convert it to a standard unified format, such as PDF, to ensure consistency in subsequent processing. If the file is encrypted, an attempt is made to decrypt it using a preset universal password library; if decryption fails, it is marked as "abnormal," the process is terminated, and an alarm is triggered. After successful decryption, the product code and report time are extracted from the report. Based on technologies such as keyword extraction, OCR, and large-scale model semantic recognition, product elements such as net asset value trends, sales rules, and product interpretations are extracted from the unstructured report information to form structured product tags. Among them, the net asset value trend refers to the price change curve of each share of the financial product over a period of time; the sales rules refer to the specific terms and regulations disclosed in the report regarding how to buy and sell the fund product, how fees are charged, and trading restrictions; the product interpretation information includes, but is not limited to, the net asset value per share at the end of the period, the cumulative net asset value per share, the net asset value growth rate for the period, the data on changes in size, a summary of the fund manager's views, the market outlook, and a list of the top ten holdings. Then, based on the extracted product elements, the system constructs structured product tags in formats such as JSON. The "product code" and "report end date" are extracted and combined to generate a globally unique identifier (ID). This unique identifier, along with the structured product tag and the financial product operation report, is then stored in a unified database. This invention collects reports from multiple data sources, ensuring data integrity and timeliness regardless of whether the report originates from external administrator emails, internal colleague uploads, or system synchronization. This eliminates omissions and achieves fully automated report collection without manual intervention, significantly improving collection efficiency.

[0025] 103. Determine the set of target users with report viewing permissions, and obtain the environmental fingerprint information of the terminal used by each target user in the target user set, wherein the environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status.

[0026] Among them, network status refers to the current communication link quality and type of the user terminal, including but not limited to real-time bandwidth, throughput, network type, network latency, packet loss rate, and signal strength; device performance parameters include but are not limited to the user terminal's hardware device's CPU model and number of cores, GPU rendering capability, total memory, current CPU utilization, remaining available memory space, battery status, screen resolution, pixel density and other screen characteristics, and local available storage space; user interface focus status refers to the user's current behavioral intent and attention distribution within the APP, including but not limited to the user's current page path, the UI component where the user's gaze or finger is currently resting, screen touch frequency, scrolling speed, whether the APP is running in the foreground or suspended in the background, whether the user has just searched for the product code, or is browsing similar competitor reports.

[0027] In this embodiment of the invention, the system reads the product code of the report to be distributed, queries the core transaction database, and extracts all user IDs currently holding shares of that product, forming a basic candidate pool. Then, it matches the user's risk tolerance rating with the product's risk level. If a user's risk rating is lower than the product requirements, or if the user's account is frozen or closed, they are removed from the candidate pool. For clients who do not hold the report but are marked as "high potential" and authorized by their financial advisor to view it, they are included in the set after secondary authentication, thus determining the target user set. The system collects the user terminal's environmental fingerprint information through a long connection channel between the client and the server. By determining the target user set with report viewing permissions, this embodiment of the invention ensures that sensitive financial reports are only distributed to compliant holders or authorized personnel, reducing the risk of information leakage.

[0028] 104. Based on the content of the financial product operation report stored in relation to the database, the financial product operation report is decomposed into multiple logical sub-streams, and a personalized streaming transmission strategy is generated for each target user based on environmental fingerprint information. The logical sub-streams are pushed to the corresponding target user's terminal based on the streaming transmission strategy. The logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream. The streaming transmission strategy defines at least one of the following for each logical sub-stream: transmission order, compression ratio, and start / stop control rules.

[0029] In this embodiment of the invention, the intelligent decomposition engine decomposes the report content into three independent logical sub-flows based on the semantic attributes, rendering dependencies, and compliance sensitivity. Each sub-flow is tagged with metadata, including tags for priority, sensitivity, and data volume. The core data flow contains the most critical text and numerical information in the report, including product code, latest unit net asset value, cumulative net asset value, current period growth rate, summary of the fund manager's core viewpoints, and risk level warnings. It has the highest priority, lower sensitivity, and smallest data volume. The visualization chart data flow includes all graphical elements in the report, such as net asset value trend charts, asset allocation pie charts, industry distribution bar charts, and graphical displays of the top ten holdings. It has a medium priority, medium-high sensitivity, and medium data volume. The full-document archive data flow refers to the original complete financial product operation report file, which has a low priority, high sensitivity, and largest data volume.

[0030] Furthermore, a streaming strategy needs to be determined. Based on the user terminal's environmental fingerprint information, a streaming strategy is formulated for each logical sub-stream, specifying the sending order, compression ratio, and start / stop control rules. The start / stop control rules dynamically determine whether to continue sending, pause sending, or interrupt the connection based on the user's real-time behavior. Finally, reports are pushed to the user terminal through a long-lived connection channel according to the streaming strategy. For example, regardless of the user's network conditions, the core data stream is pushed immediately, and a visualization chart data stream is pushed according to the compression ratio in the strategy. For example, if the strategy determines a weak network, a blurry placeholder image is pushed first, and a high-resolution detailed image is pushed as needed when the user clicks to zoom in or the network improves. If the strategy determines that the user has no intention of viewing the chart, such as when the user quickly swipes to skip, the remaining portion of the visualization chart data stream is postponed or canceled according to the start / stop rules. After the core data stream and visualization chart data stream have been transmitted and the user is still on the page, a full-document file data stream is pushed in the background with low priority, utilizing idle bandwidth. This invention significantly reduces loading failures caused by network fluctuations or device incompatibility by dynamically adjusting the transmission strategy according to the real-time environment. For high-concurrency scenarios, this invention splits the report into streams, allowing caching and acceleration only for frequently accessed "core data," reducing peak bandwidth pressure on large files. Through the streaming transmission strategy, authentication and distribution can be achieved immediately after report parsing, significantly improving timeliness compared to the traditional manual distribution model.

[0031] Furthermore, in order to record the processing progress of the report and identify anomalies, a processing log needs to be constructed. Based on this, the method includes: generating a processing log with transaction tracking identifiers based on the processing process of the financial product operation report, wherein the processing process includes at least one of the following: a report receiving rule configuration process, a collection process, a parsing process, a structured processing process, a storage process, a logical sub-stream decomposition process, and a logical sub-stream push process; identifying abnormal events in the processing process of the financial product operation report based on the processing log, performing root cause analysis on the abnormal events, determining the anomaly type of the abnormal events based on the root cause analysis results, determining an anomaly notification strategy based on the anomaly type, and pushing the abnormal events based on the anomaly notification strategy.

[0032] Specifically, upon receiving any signal that triggers report processing (such as the start of a scheduled task, the arrival of a new email, or manual upload), the system immediately generates a globally unique transaction tracking identifier. Log points are embedded in key processes such as rule configuration, report collection, parsing, structured processing, storage, logical sub-stream decomposition, and logical sub-stream push. Based on these log points, the system obtains the following information: rule loading status and regular expression compilation results during rule configuration; source type, report file hash value, download time, and network status during report collection; extracted tag information during parsing and structured processing; database write status, object storage path, number of sub-stream splits, and tag generation results during storage and decomposition; and the number of target users, number of successful transmissions, list of failed user IDs, and average transmission latency during push. All of this information is recorded in the processing log. For each processing step, a processing baseline is constructed based on historical data. Abnormal events are identified based on the deviation between the processing step and the baseline in the logs. For example, the system learns from historical data from the same period, such as the processing time baseline over the past 30 days. If the parsing time of a report exceeds the baseline by three standard deviations, it is judged as a "performance degradation anomaly in report parsing." Root cause analysis is then performed on the abnormal events. Based on the root cause analysis results, the abnormal events are categorized into their respective anomaly types. These anomaly types include, but are not limited to, system-level failures, business anomalies, performance fluctuation anomalies, and data quality anomalies. For example, if the root cause analysis results in a connection refusal to connect to a specific fund company's email server, parsing failure due to a report template change, or data validation failure, then the corresponding anomaly type is determined to be a business anomaly.

[0033] Furthermore, different anomaly notification strategies are constructed for different anomaly types. Specifically, the anomaly level corresponding to the anomaly type can be determined, and the anomaly notification strategy can be determined based on the anomaly level. For example, for low-level anomalies, anomaly event notification is pushed to maintenance personnel's terminals via robot push. The message card pushed by the robot not only contains a text description of the anomaly event but also embeds an interactive operation button. Clicking the operation button allows viewing the entire link log. For high-level anomalies, in addition to pushing the anomaly event through communication media such as email, outbound calls are also made to maintenance personnel via telephone or other outbound calling devices. At the same time, the corresponding handling strategy for the anomaly event can be determined, and the handling strategy is also attached when pushing the message to the maintenance personnel's terminals, so that maintenance personnel can quickly and accurately handle the anomaly event. This embodiment of the invention can connect logs scattered in all stages such as configuration, collection, parsing, storage, disassembly, and push into a complete call chain through a transaction tracking and identification system. Once an anomaly occurs, the system automatically traces back the entire process of the ID and instantly locates the fault location. By configuring different notification strategies for different anomaly types, the notification effect can be improved.

[0034] According to the present invention, a method for processing financial product operation reports, compared with the current method of manually processing financial product operation reports, allows for the configuration of report receiving rules to simultaneously accommodate multiple operation report sources. This enables automatic monitoring of multiple channels, eliminating the need for manual login to the system to collect reports one by one, thus completely solving the problems of low efficiency and easy omissions in manual collection. Furthermore, by using regular expressions to accurately match key information in filenames or email titles, it effectively filters spam and irrelevant files, ensuring data purity from the source. Different fund managers send reports at different frequencies, and the configurable synchronization frequency allows the system to flexibly adapt to various business scenarios, neither excessively consuming resources nor missing urgent reports. By aggregating reports scattered from different sources to a unified platform, data fragmentation is eliminated, facilitating centralized processing. The system manages and associates data through a unique product identifier, effectively preventing duplicate data from multiple uploads or receptions of the same report, ensuring the cleanliness and uniqueness of the database. By defining the target user set with report viewing permissions, it ensures that sensitive financial reports are only distributed to compliant holders or authorized personnel, reducing the risk of information leakage. By dynamically adjusting the transmission strategy according to the real-time environment, it significantly reduces loading failures caused by network fluctuations or device incompatibility. For high-concurrency scenarios, this invention splits the report into streams, allowing caching and acceleration only for frequently accessed "core data," reducing peak bandwidth pressure on large files. Through the streaming transmission strategy, it enables immediate authentication and distribution after report parsing, significantly improving timeliness compared to the traditional manual distribution model.

[0035] Furthermore, to better illustrate the above process of processing financial product operation reports, as a refinement and extension of the above embodiments, this invention provides another method for processing financial product operation reports, such as... Figure 2 As shown, the method includes: 201. In response to the processing signal of the financial product operation report, configure the report receiving rules, wherein the report receiving rules include at least one of the following: report source address, report identification regular expression, and report synchronization frequency.

[0036] Specifically, users configure report collection through a specific interactive interface of the system. Configuration items include the administrator's email address, parsing rule regular expression, data synchronization frequency, report type label, etc.

[0037] 202. Based on the report receiving rules, automatically collect financial product operation reports from multiple preset sources, and uniformly parse and structure the collected financial product operation reports to generate structured product tags containing unique product identifiers. Store the financial product operation reports and structured product tags together. The preset sources include at least one of email, internal marketing system, and file upload interface.

[0038] In this embodiment of the invention, the following methods are used: Email parsing and collection by the manager: Operational reports are collected by parsing attachments sent by the manager via email. Internal material system collection: Files related to operational reports are automatically retrieved from the internal employee marketing system. Manual upload: For a small number of reports that the system cannot automatically retrieve, manual upload is provided. In this way, financial product operational reports are collected from multiple preset sources. Furthermore, for each product report, its product code, report time, and other information are determined to form a unique identifier for storage in the database. Through keyword extraction, OCR, and large-scale semantic recognition technologies, product elements such as net asset value trends, sales rules, and product interpretations are extracted from unstructured report information to form structured product tags. Furthermore, to improve the accuracy of product tag extraction, the product tags also need to be verified. Based on this, the method includes: retrieving standard benchmark data corresponding to the financial product operation report from a preset standardized financial database using the product code as an index key; wherein the standard benchmark data includes standard product name, standard report time, standard net asset value trend, standard sales rules, and standard product interpretation; verifying the accuracy of the structured product tags using the standard benchmark data; generating a data correction strategy based on the accuracy verification results; and correcting the structured product tags using the data correction strategy.

[0039] The structured product tags include at least one of the following: product code, product name, report time, net asset value trend, sales rules, and product interpretation; the standard benchmark data includes the standard product name, standard report time, standard net asset value trend, standard sales rules, and standard product interpretation of the financial product operation report; and the standard benchmark data of various financial product operation reports are stored in a pre-set standardized financial database.

[0040] Specifically, the similarity between the product's structured tag and the corresponding standard benchmark data is calculated. If the similarity is greater than a preset threshold set according to actual needs, the product's structured tag is deemed to have passed verification; if the similarity is less than or equal to the preset threshold, the product's structured tag is deemed to have failed verification. For product structured tags that fail verification, a correction strategy is determined based on the standard benchmark data that it supports, and the correction strategy is used to correct the product's structured tag. For example, the corresponding field in the standard benchmark data can be directly used to cover the erroneous field in the product's structured tag. Then, the report is categorized according to the unique identifier, and the unique identifier, the corrected structured product tag, and the report are associated and stored in the cloud based on the categorization category. This embodiment of the invention improves data quality by verifying product tags.

[0041] 203. Determine the set of target users with report viewing permissions, and obtain the environmental fingerprint information of the terminal used by each target user in the target user set, wherein the environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status.

[0042] In this embodiment of the invention, to ensure data security, it is first necessary to determine the target user set with report viewing permissions. Based on this, step 203 specifically includes: determining multi-dimensional association information between multiple candidate users and the financial product operation report, user characteristic information of the candidate users, and report characteristic information of the financial product operation report. The multi-dimensional association information includes the candidate users' holding characteristics, behavioral characteristics, and implicit associations with the financial products in the financial product operation report; based on the multi-dimensional association information, the user characteristic information, and the report characteristic information, generating a permission score for each candidate user, and matching the target user set with report viewing permissions among each candidate user based on the permission score.

[0043] The information includes: holding characteristics, holding percentage, holding period, profit / loss status, and trading frequency; behavioral characteristics, browsing frequency, dwell time, search history, and consultation history of the candidate user regarding the products in the report or similar products; implicit relationships, whether the user has a potential purchase intention for the products in the report; user characteristics, including risk tolerance level, age, gender, region, occupation, and interests; and report characteristics, including product risk level, report sensitivity, and report content type.

[0044] Specifically, candidate users can be registered users of the company to which the financial product delivery report belongs. The process of determining the permission score for each candidate user is as follows: First, determine the association feature vector corresponding to the multi-dimensional association information, the user feature vector corresponding to the user feature information, and the report feature vector corresponding to the report feature information. Second, perform feature-level cross processing on the association feature vector, the user feature vector, and the report feature vector to obtain a feature cross vector. Third, perform element-level cross processing on the association feature vector, the user feature vector, and the report feature vector to obtain an element cross vector. Fourth, perform low-order cross processing on the association feature vector, the user feature vector, and the report feature vector to obtain a low-order cross vector. Fifth, transform the feature cross vector, the element cross vector, and the low-order cross vector to obtain a permission cross feature vector. Finally, input the permission cross feature vector into a preset scoring model for scoring to obtain the permission score for the corresponding candidate user.

[0045] In another embodiment of the present invention, in order to improve the scoring accuracy of the preset scoring model, it is first necessary to train and construct the preset scoring model. Based on this, the method includes: constructing a preset initial scoring model; obtaining a sample dataset, wherein the sample dataset includes multi-dimensional correlation information between sample users with permission scoring labels and sample financial products, the user feature information, and the report feature information; dividing the sample dataset into a training set and a test set, using the training set to train the preset initial scoring model, and using the test set to test the trained preset initial scoring model, and finally using the trained preset initial scoring model that meets the test conditions as the preset scoring model. Specifically, in the model training process, the preset initial scoring model is first constructed, and then the sample dataset is obtained. Ensure that the dataset contains all necessary files. Convert the data into a format that the preset initial scoring model can understand, and finally train and test the model. Specifically, the dataset can be divided first: using random or specific strategies (such as stratified sampling) to divide the sample dataset into a training set and a test set. Then, the model is trained using the training set, and the trained model is tested using the test set to evaluate its performance on unseen data. Calculate and record metrics such as precision and recall on the test set. If the model performance does not meet the requirements, return to the training phase for further iterations or adjustments. This process yields a pre-defined scoring model that meets the requirements. This pre-defined scoring model includes an input layer, hidden layers, and a scoring layer.

[0046] Furthermore, feature extraction models, such as CNN models, are used to extract associated feature vectors, user feature vectors, and report feature vectors respectively. If the associated feature vector is (a1, a2), the user feature vector is (b1, b2), and the report feature vector is (c1, c2), the specific cross-processing method includes: performing feature-level cross-processing between different feature vectors, i.e., performing a Hadamard product on all elements of the vectors, followed by a convolution transformation under certain weights w1, to obtain the feature cross vector f(w1×(a1×b1×c1, a2×b2×c2)); simultaneously, performing element-level cross-processing on all feature vector data. That is, after performing a Hadamard product on each element of the vectors, different weight values ​​w2 and w3 are assigned to the result of each product, and then a linear transformation is performed to obtain the element-wise cross vector f(w2×a1×b1×c1, w3×a2×b2×c2). Furthermore, all feature vectors undergo low-order cross processing, and the result of the cross processing is assigned a weight coefficient w4, followed by a linear transformation to obtain the low-order cross vector f(w4(a1,a2,b1,b2,c1,c2)). Finally, the above feature cross vectors, element-wise cross vectors, and low-order cross vectors are combined, such as by horizontal concatenation, to obtain the permissioned cross feature vector. It should be noted that the above examples are merely illustrative and do not limit the embodiments of this application. Therefore, by cross-processing the associated feature vector, user feature vector, and report feature vector, different features can be automatically or explicitly combined to generate new feature combinations. These combined features may contain complex nonlinear relationships between the original features, enabling the model to capture more refined and richer information in the data. This means it can fully utilize the relationships between various data points, extract more latent features, and simultaneously handle both high-order and low-order processing, making data utilization more efficient and resulting in more accurate scoring results that meet the needs of practical applications. Finally, the permission cross-feature vector is directly input into the input layer of the preset scoring model. The input layer then transmits the permission cross-feature vector to the hidden layer, where feature enhancement processing is performed. The scoring layer then scores the output features of the hidden layer to obtain the permission scores of candidate users. Finally, candidate users with high permission scores are selected to form the target user set.

[0047] Furthermore, the environmental fingerprint information of the terminal device used by each target user in the target user set is obtained.

[0048] 204. Extract key numerical indicators and text summaries from financial product operation reports, and encapsulate the key numerical indicators and text summaries into a core data stream.

[0049] Specifically, the parsed structured reports are first subjected to in-depth screening. Using a predefined dictionary of key financial indicators and a large language model summarization algorithm, key numerical indicators (including product code, unit net asset value, cumulative net asset value, current period growth rate, asset size, and the percentage of the top ten holdings) and text summaries (including the fund manager's core views, key market outlook, and important risk warnings) are precisely extracted from the reports. Subsequently, this high-value, low-volume data is serialized into lightweight JSON or XML format data packets, marked with the highest priority (P0), and encapsulated to generate the core data stream. This stream has an extremely small data volume, dedicated to millisecond-level initial screen rendering on user terminals.

[0050] 205. Identify the charts, curves, and image elements in the financial product operation report, convert the charts, curves, and image elements into vector data respectively, and encapsulate the vector data into a visual chart data stream.

[0051] Specifically, computer vision technology is used to locate all charts, curves, and image elements in the report (such as net asset value trend line charts, asset allocation pie charts, industry distribution bar charts, and scale change trend charts). For identified bitmaps or static images, the system calls a vector conversion engine to redraw or convert them into vector data formats (such as SVG, PDF vector instruction sets, or Canvas drawing instructions), ensuring that the images remain clear at any scaling ratio and that the file size is controllable. The system indexes the converted vector data by chart ID, packages it, marks it as medium priority (P1), and encapsulates it to generate a visual chart data stream. This stream supports on-demand loading and progressive rendering.

[0052] 206. Encapsulate the entire financial product operation report into a full-document archive data stream, and treat the core data stream, visualization chart data stream, and full-document archive data stream as multiple logical sub-streams.

[0053] Specifically, the entire financial product operation report, after being cleaned and standardized, is packaged into a complete binary file or high-resolution document stream, following the original layout order or logical chapter order. This data packet is marked as the lowest priority (P2) and serves as the full-document file data stream, primarily for in-depth user review, offline archiving, and compliance auditing. It is typically transmitted using idle background bandwidth after the core data and visualization content have been loaded. This method effectively breaks down the report into multiple logical sub-streams.

[0054] 207. Generate a personalized streaming strategy for each target user based on environmental fingerprint information, and push logical sub-streams to the corresponding target user's terminal based on the streaming strategy. The logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream. The streaming strategy defines at least one of the following for each logical sub-stream: transmission order, compression ratio, and start / stop control rules.

[0055] In this embodiment of the invention, in order to push logical sub-streams, it is first necessary to determine the streaming transmission strategy of the logical sub-streams. Based on this, step 207 specifically includes: if it is detected that the user terminal is in a weak network environment and the user interface focus is on the report details page, the streaming transmission strategy is set to transmit the core data stream at full speed only, the visualization chart data stream is downgraded to low-resolution thumbnail transmission, and the transmission of the full document file data stream is paused; if it is detected that the user terminal is in a high bandwidth environment and the user interface focus is on the message list, the streaming transmission strategy is set to silently preload the core data stream, visualization chart stream, and full document file data stream to the local cache in the background; if it is detected that the user terminal's battery level is lower than a preset threshold, the streaming transmission strategy is set to limit the transmission frame rate or resolution of the visualization chart data stream.

[0056] Specifically, based on environmental fingerprint information, when the terminal is detected to be in a weak network environment and the user interface focus is on the report details page, the strategy is set to "prioritize core data, downgrade visual data, and stop archiving." This means: high-priority core data streams are transmitted at full speed to ensure immediate display on the first screen; the visual chart data stream is downgraded to low-resolution thumbnails for transmission to save bandwidth; and the transmission of the full-document file data stream is paused to avoid blocking the loading of critical information. When the terminal is detected to be in a high-bandwidth environment (such as Wi-Fi) and the user interface focus is on the message list (not the current reading page), the strategy is set to "silent preloading in the background." The system uses idle bandwidth in the background to preload the core data stream, visual chart data stream, and full-document file data stream sequentially and completely to the local cache, ensuring that the user can quickly view the report content when clicking to enter the report. When the remaining battery power of the terminal is detected to be lower than the preset threshold (the preset threshold is set according to actual needs), the strategy is set to "power consumption priority". The transmission specifications of the visualization chart data stream are automatically limited, including reducing image resolution, reducing animation frame rate or simplifying vector complexity, so as to reduce the energy consumption of terminal screen rendering and data decoding, extend device battery life, and ensure the normal reading of core text information.

[0057] Furthermore, each logical sub-stream is pushed according to the transmission order, compression ratio, and start / stop control rules set in the streaming transmission strategy. For example, priority-based push: first, high-priority core data streams are pushed in milliseconds to achieve instant rendering of the first screen; then, medium-priority visualization chart data streams are pushed, supporting progressive loading; finally, low-priority full-text file data streams are pushed in the background for in-depth viewing. Environment-adaptive compression: the sub-stream form is dynamically adjusted according to the compression ratio set in the strategy. For example, in weak network or low-configuration devices, visualization streams are automatically compressed at a high ratio or converted to thumbnails, and high compression algorithms are used for full-text streams. In high-bandwidth environments, lossless or low-compression ratio transmission is used to ensure a high-definition experience. Interaction-based dynamic start / stop: the start / stop control rules in the strategy are executed. For example, when a user enters the details page, the corresponding sub-stream transmission is started immediately; if the user quickly scrolls over, exits the page, or there is a sudden network fluctuation, the transmission of subsequent low-priority sub-streams (such as charts and full text) is immediately paused or terminated, and breakpoint resume is supported when the user returns, avoiding resource waste.

[0058] Furthermore, without real-time adjustments to the report's push strategy, after pushing logical sub-streams to the corresponding target user's terminal based on the streaming strategy, the method further includes: deploying a behavior probe on the target user's terminal; using the behavior probe to capture interactive events of the target user reading the financial product operation report in real time; wherein the interactive events include at least one of page scrolling speed, page scrolling acceleration, chart zoom level changes, selected text content, copied text content, cursor or touch hover trajectory, and cursor or touch dwell time; identifying the target user's current reading behavior pattern based on the interactive events; wherein the current reading behavior pattern includes one of in-depth analysis mode, skimming and filtering mode, transaction intent mode, and confusion and hesitation mode; generating a dynamic content heat map based on each target user's current reading behavior pattern; determining high-attention content segments in the financial product operation report based on the dynamic content heat map; and adjusting the streaming strategy for subsequent users reading the financial product operation report based on the high-attention content segments; wherein adjusting the streaming strategy includes at least one of automatically rearranging chapter order, preloading high-attention data blocks, and automatically associating interpretation auxiliary information with high-difficulty content segments.

[0059] Specifically, for example: if a user's interaction events while reading a financial product operation report include repeatedly zooming in and out of specific charts or hovering for extended periods, their current reading behavior pattern is determined to be an in-depth analysis mode; if the user's interaction events include high-speed mouse scrolling or skipping without interaction, their current reading behavior pattern is determined to be a skimming and filtering mode; if the user's interaction events include copying fees, redemption rules, or key performance data, their current reading behavior pattern is determined to be a trading intent mode; if the user's interaction events include repeatedly scrolling back and forth in a complex paragraph or remaining motionless for extended periods, their current reading behavior pattern is determined to be a confused and hesitant mode. Real-time aggregation of all users' behavior pattern data generates a dynamic content heatmap, accurately identifying "high-attention content segments" (such as frequently studied position analyses) and "high-difficulty / confused content segments" (such as terminology sections that cause excessively long dwell times) within the report. Then, based on the heatmap, the streaming strategy for subsequent users accessing the report is optimized in real time. This includes: automatic reordering: such as prioritizing logical sub-streams corresponding to highly popular chapters and placing them at the top of the transmission sequence to ensure users see the core focus on their first screen; intelligent preloading: such as pre-transmitting highly popular data blocks (e.g., a frequently magnified curve chart) to the local cache via the background stream channel before the user clicks; and auxiliary information association: for content fragments marked as "confused," auxiliary information streams such as popular interpretations, glossaries, or financial advisor comments are dynamically embedded and prioritized during the transmission of the corresponding sub-stream. By recognizing the current user's reading behavior patterns, this invention can understand the user's focus and level of need for the report content. Based on this information, the transmission strategy can be adjusted in real time, helping subsequent readers better understand complex content and enabling them to quickly obtain the content they need, thus meeting their personalized reading needs.

[0060] According to another method for processing financial product operation reports provided by this invention, compared with the current method of manually processing financial product operation reports, this invention, by configuring report receiving rules to be compatible with multiple operation report sources simultaneously, can automatically monitor multiple channels without requiring manual login to the system to collect reports one by one, completely solving the problems of low efficiency and easy omissions in manual collection. Furthermore, by using regular expressions to accurately match key information in filenames or email titles, it effectively filters spam and irrelevant files, ensuring data purity from the source. Different fund managers send reports at different frequencies, and the configurable synchronization frequency allows the system to flexibly adapt to various business scenarios, neither excessively consuming resources nor missing urgent reports. By aggregating reports scattered from different sources to a unified platform, data fragmentation is eliminated, facilitating data aggregation. The system manages reports centrally and uses a unique product identifier for associated storage, effectively preventing duplicate data from multiple uploads or receptions of the same report, ensuring the cleanliness and uniqueness of the database. By defining the target user set with report viewing permissions, it ensures that sensitive financial reports are only distributed to compliant holders or authorized personnel, reducing the risk of information leakage. By dynamically adjusting the transmission strategy according to the real-time environment, it significantly reduces loading failures caused by network fluctuations or device incompatibility. For high-concurrency scenarios, this invention splits reports into streams, allowing caching and acceleration only for frequently accessed "core data," reducing peak bandwidth pressure on large files. Through a streaming transmission strategy, it enables immediate authentication and distribution after report parsing, significantly improving timeliness compared to the traditional manual distribution model.

[0061] Furthermore, as Figure 1 In a specific implementation, embodiments of the present invention provide a processing apparatus for financial product operation reports, such as... Figure 3 As shown, the device includes: a configuration unit 31, a storage unit 32, an acquisition unit 33, and an information push unit 34.

[0062] The configuration unit 31 can be used to configure report receiving rules in response to the processing signal of financial product operation report, wherein the report receiving rules include at least one of report source address, report identification regular expression, and report synchronization frequency.

[0063] The storage unit 32 can be used to automatically collect financial product operation reports from multiple preset sources based on the report receiving rules, and perform unified parsing and structuring processing on the collected financial product operation reports to generate structured product tags containing unique product identifiers. The financial product operation reports are then associated with and stored with the structured product tags. The preset sources include at least one of email, internal marketing systems, and file upload interfaces.

[0064] The acquisition unit 33 can be used to determine a set of target users with report viewing permissions and acquire environmental fingerprint information of the terminal used by each target user in the set of target users. The environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status.

[0065] The information push unit 34 can be used to decompose the financial product operation report into multiple logical sub-streams based on the content of the associated stored financial product operation report, and generate a personalized streaming transmission strategy for each target user based on the environmental fingerprint information. Based on the streaming transmission strategy, the logical sub-streams are pushed to the corresponding target user's terminal. The logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream. The streaming transmission strategy defines at least one of the following: transmission order, compression ratio, and start / stop control rules for each logical sub-stream.

[0066] In specific application scenarios, in order to break down financial product operation reports into multiple logical sub-streams, such as... Figure 4 As shown, the information push unit 34 includes an extraction module 341 and an encapsulation module 342.

[0067] The extraction module 341 can be used to extract key numerical indicators and text summaries from the financial product operation report, and encapsulate the key numerical indicators and text summaries into a core data stream.

[0068] The encapsulation module 342 can be used to identify the charts, curves, and image elements in the financial product operation report, convert the charts, curves, and image elements into vector data respectively, and encapsulate the vector data into a visual chart data stream.

[0069] The encapsulation module 342 can also be used to encapsulate the entire financial product operation report into a full-document file data stream.

[0070] In specific application scenarios, in order to determine the streaming transmission strategy, the information push unit 34 also includes a setting module 343.

[0071] The setting module 343 can be configured to, if it is detected that the user terminal is in a weak network environment and the user interface focus is on the report details page, set the streaming transmission strategy to transmit only the core data stream at full speed, downgrade the visualization chart data stream to low-resolution thumbnail transmission, and pause the transmission of the full document file data stream; if it is detected that the user terminal is in a high bandwidth environment and the user interface focus is on the message list, set the streaming transmission strategy to silently preload the core data stream, visualization chart stream, and full document file data stream to the local cache in the background; if it is detected that the user terminal's battery level is lower than a preset threshold, set the streaming transmission strategy to limit the transmission frame rate or resolution of the visualization chart data stream.

[0072] In specific application scenarios, in order to determine the target user set with report viewing permissions, the acquisition unit 33 includes a determination module 331 and a permission scoring module 332.

[0073] The determining module 331 can be used to determine multi-dimensional correlation information between multiple candidate users and the financial product operation report, user characteristic information of the candidate users, and report characteristic information of the financial product operation report. The multi-dimensional correlation information includes the candidate users' holding characteristics information, behavioral characteristics information, and implicit correlation relationships of the financial products in the financial product operation report.

[0074] The permission scoring module 332 can be used to generate a permission score for each candidate user based on the multi-dimensional association information, the user feature information, and the report feature information, and to match a set of target users with report viewing permissions among each candidate user based on the permission score.

[0075] In specific application scenarios, in order to identify and push abnormal events during the report processing, the device also includes an abnormal push unit 35.

[0076] The anomaly push unit 35 can be used to generate a processing log with a transaction tracking identifier based on the processing process of the financial product operation report. The processing process includes at least one of the following: a report receiving rule configuration process, a collection process, a parsing process, a structured processing process, a storage process, a logical sub-stream decomposition process, and a logical sub-stream push process. Based on the processing log, anomaly events in the financial product operation report processing process are identified, and root cause analysis is performed on the anomaly events. Based on the root cause analysis results, the anomaly type of the anomaly event is determined, an anomaly notification strategy is determined based on the anomaly type, and the anomaly event is pushed based on the anomaly notification strategy.

[0077] In specific application scenarios, in order to adjust the streaming transmission strategy when users read financial product operation reports, the device also includes a distribution strategy adjustment unit 36.

[0078] The distribution strategy adjustment unit 36 ​​can be used to deploy a behavior probe on the target user's terminal, and use the behavior probe to capture interactive events of the target user reading the financial product operation report in real time. These interactive events include at least one of the following: page scrolling speed, page scrolling acceleration, chart zoom level changes, selected text content, copied text content, cursor or touch hover trajectory, and cursor or touch dwell time. Based on these interactive events, the unit identifies the target user's current reading behavior pattern, which includes one of the following: in-depth analysis mode, skimming and filtering mode, transaction intent mode, and confusion / hesitation mode. Based on each target user's current reading behavior pattern, a dynamic content heat map is generated. Based on the dynamic content heat map, high-attention content segments in the financial product operation report are determined. Based on these high-attention content segments, the streaming transmission strategy for subsequent users reading the financial product operation report is adjusted. The adjusted streaming transmission strategy includes at least one of the following: automatically rearranging chapter order, preloading high-attention data blocks, and automatically associating interpretation auxiliary information with high-difficulty content segments.

[0079] In specific application scenarios, the structured product tags include at least one of product code, product name, report time, net value trend, sales rules, and product interpretation; in order to verify the structured product tags, the device also includes a verification unit 37.

[0080] The verification unit 37 can be used to retrieve the standard benchmark data corresponding to the financial product operation report in a preset standardized financial database using the product code as the index key. The standard benchmark data includes the standard product name, standard report time, standard net value trend, standard sales rules, and standard product interpretation. The unit can also use the standard benchmark data to verify the accuracy of the structured product label, generate a data correction strategy based on the accuracy verification result, and use the data correction strategy to correct the structured product label.

[0081] It should be noted that other corresponding descriptions of the functional modules involved in the financial product operation report processing device provided in this embodiment of the invention can be found in the following references. Figure 1 The corresponding description of the method shown will not be repeated here.

[0082] Based on the above, Figure 1The method shown in the invention, correspondingly, also provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the program performs the following steps: responding to a processing signal for a financial product operation report, configuring report receiving rules, wherein the report receiving rules include at least one of a report source address, a report identification regular expression, and a report synchronization frequency; based on the report receiving rules, automatically collecting financial product operation reports from multiple preset sources, uniformly parsing and structuring the collected financial product operation reports to generate a structured product tag containing a unique product identifier, and storing the financial product operation report associated with the structured product tag, wherein the preset sources include at least one of email, an internal marketing system, and a file upload interface; determining the user with the right to view the report. The system identifies a target user set for permissions and obtains the environmental fingerprint information of the terminal used by each target user in the target user set. The environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status. Based on the content of the financial product operation report stored in association, the financial product operation report is decomposed into multiple logical sub-streams. A personalized streaming transmission strategy is generated for each target user based on the environmental fingerprint information. The logical sub-streams are pushed to the terminal used by the corresponding target user based on the streaming transmission strategy. The logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream. The streaming transmission strategy defines at least one of the transmission order, compression ratio, and start / stop control rules for each logical sub-stream.

[0083] Based on the above, Figure 1 The method shown and as Figure 3 The embodiment of the device shown in the invention also provides a physical structure diagram of a computer device, such as... Figure 5As shown, the computer device includes: a processor 41, a memory 42, and a computer program stored on the memory 42 and executable on the processor. Both the memory 42 and the processor 41 are mounted on a bus 43. When the processor 41 executes the program, it performs the following steps: responding to a processing signal for a financial product operation report, configuring report receiving rules, wherein the report receiving rules include at least one of a report source address, a report identification regular expression, and a report synchronization frequency; based on the report receiving rules, automatically collecting financial product operation reports from multiple preset sources, uniformly parsing and structuring the collected financial product operation reports to generate a structured product tag containing a unique product identifier, and storing the financial product operation report associated with the structured product tag. The preset sources include email, internal marketing systems, and file uploads. At least one of the following interfaces; determining a set of target users with report viewing permissions, and obtaining environmental fingerprint information of the terminal used by each target user in the target user set, wherein the environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status; based on the content of the financial product operation report stored in association, decomposing the financial product operation report into multiple logical sub-streams, and generating a personalized streaming transmission strategy for each target user based on the environmental fingerprint information, and pushing the logical sub-streams to the terminal used by the corresponding target user based on the streaming transmission strategy, wherein the logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream, and the streaming transmission strategy defines at least one of the transmission order, compression ratio, and start / stop control rules for each of the logical sub-streams.

[0084] Through the technical solution of this invention, the system can simultaneously accommodate multiple operational report sources by configuring report reception rules. It can automatically monitor multiple channels without requiring manual login to each system for data collection, completely solving the problems of low efficiency and easy omissions associated with manual collection. Furthermore, it utilizes regular expressions to accurately match key information in filenames or email titles, effectively filtering spam and irrelevant files, ensuring data purity from the source. Different fund managers send reports at different frequencies; the configurable synchronization frequency allows the system to flexibly adapt to various business scenarios, neither excessively consuming resources nor missing urgent reports. By aggregating reports scattered from different sources to a unified platform, data fragmentation is eliminated, facilitating centralized management, and the reports are associated and stored using a "product unique identifier." This invention effectively avoids duplicate data caused by multiple uploads or receptions of the same report, ensuring the cleanliness and uniqueness of the database. By defining the target user set with report viewing permissions, it ensures that sensitive financial reports are only distributed to compliant holders or authorized personnel, reducing the risk of information leakage. By dynamically adjusting the transmission strategy according to the real-time environment, it significantly reduces loading failures caused by network fluctuations or device incompatibility. For high-concurrency scenarios, this invention splits the report into streams, allowing caching and acceleration only for frequently accessed "core data," reducing peak bandwidth pressure on large files. Through the streaming transmission strategy, it enables immediate authentication and distribution after report parsing, significantly improving timeliness compared to the traditional manual distribution model.

[0085] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0086] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for processing financial product operation reports, characterized in that, include: In response to the processing signal of the financial product operation report, a report receiving rule is configured, wherein the report receiving rule includes at least one of the following: report source address, report identification regular expression, and report synchronization frequency; Based on the report receiving rules, financial product operation reports are automatically collected from multiple preset sources. The collected financial product operation reports are uniformly parsed and structured to generate structured product tags containing unique product identifiers. The financial product operation reports are associated with and stored with the structured product tags. The preset sources include at least one of email, internal marketing system, and file upload interface. A set of target users with report viewing permissions is determined, and environmental fingerprint information of the terminal used by each target user in the set of target users is obtained, wherein the environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status; Based on the content of the financial product operation report stored in the association, the financial product operation report is decomposed into multiple logical sub-streams, and a personalized streaming transmission strategy is generated for each target user based on the environmental fingerprint information. The logical sub-streams are then pushed to the corresponding target user's terminal based on the streaming transmission strategy. The logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream. The streaming transmission strategy defines at least one of the following for each logical sub-stream: transmission order, compression ratio, and start / stop control rules.

2. The method according to claim 1, characterized in that, Based on the content of the financial product operation report stored in the association, the financial product operation report is decomposed into multiple logical sub-streams, including: Extract the key numerical indicators and text summaries from the financial product operation report, and encapsulate the key numerical indicators and text summaries into a core data stream; The system identifies the charts, curves, and image elements in the financial product operation report, converts the charts, curves, and image elements into vector data, and encapsulates the vector data into a visual chart data stream. The entire financial product operation report is packaged into a full-document archive data stream.

3. The method according to claim 1, characterized in that, Based on the environmental fingerprint information, a personalized streaming strategy is generated for each target user, including: If it is detected that the user terminal is in a weak network environment and the user interface focus is on the report details page, the streaming strategy is set to transmit the core data stream at full speed only, the visualization chart data stream is downgraded to low-resolution thumbnail transmission, and the transmission of the full document file data stream is suspended. If it is detected that the terminal is in a high-bandwidth environment and the user interface focus is on the message list, the streaming strategy is set to silently preload the core data stream, visual icon stream, and full document file data stream to the local cache in the background; If the battery level of the terminal is detected to be below a preset threshold, the streaming policy is set to limit the frame rate or resolution of the visualization chart data stream.

4. The method according to claim 1, characterized in that, Identify the target set of users with report viewing permissions, including: The system identifies multi-dimensional correlation information between multiple candidate users and the financial product operation report, user characteristic information of the candidate users, and report characteristic information of the financial product operation report. The multi-dimensional correlation information includes the candidate users' holding characteristics, behavioral characteristics, and implicit correlations with the financial products in the financial product operation report. Based on the multi-dimensional association information, the user feature information, and the report feature information, a permission score is generated for each candidate user, and a set of target users with report viewing permissions is matched among each candidate user based on the permission score.

5. The method according to claim 1, characterized in that, The method further includes: The processing of financial product operation reports generates processing logs with transaction tracking identifiers. The processing process includes at least one of the following: configuration process for report receiving rules, collection process, parsing process, structured processing process, storage process, logical sub-stream decomposition process, and logical sub-stream push process. Based on the processing logs, abnormal events in the financial product operation report processing process are identified, and root cause analysis is performed on the abnormal events. Based on the root cause analysis results, the abnormal type of the abnormal event is determined, an abnormal notification strategy is determined based on the abnormal type, and the abnormal event is pushed based on the abnormal notification strategy.

6. The method according to claim 1, characterized in that, After pushing the logical sub-stream to the terminal of the corresponding target user based on the streaming strategy, the method further includes: A behavior probe is deployed on the target user's terminal to capture interaction events in real time when the target user reads the financial product operation report. The interaction events include at least one of the following: page scrolling speed, page scrolling acceleration, chart zoom level change, text selection, text copying, cursor or touch hover trajectory, and cursor or touch dwell time. The target user's current reading behavior pattern is identified based on the interaction event, wherein the current reading behavior pattern includes one of the following: in-depth analysis mode, skimming and filtering mode, transaction intention mode, and confusion and hesitation mode. A dynamic content heat map is generated based on the current reading behavior pattern of each target user. High-attention content segments in the financial product operation report are determined based on the dynamic content heat map. The streaming transmission strategy for subsequent users reading the financial product operation report is adjusted based on the high-attention content segments. The adjustment of the streaming transmission strategy includes at least one of the following: automatically rearranging the chapter order, preloading high-attention data blocks, and automatically associating interpretation auxiliary information with high-difficulty content segments.

7. The method according to claim 1, characterized in that, The structured product tags include at least one of the following: product code, product name, report time, net asset value trend, sales rules, and product interpretation. Before storing the financial product operation report in association with the structured product tag, the method further includes: Using the product code as an index key, retrieve the standard benchmark data corresponding to the financial product operation report in a preset standardized financial database. The standard benchmark data includes the standard product name, standard report time, standard net asset value trend, standard sales rules, and standard product interpretation. The structured product label is verified for accuracy using the standard benchmark data. A data correction strategy is generated based on the accuracy verification results, and the structured product label is corrected using the data correction strategy.

8. A processing device for financial product operation reports, characterized in that, include: A configuration unit is configured to configure report receiving rules in response to a processing signal for a financial product operation report, wherein the report receiving rules include at least one of a report source address, a report identification regular expression, and a report synchronization frequency; The storage unit is used to automatically collect financial product operation reports from multiple preset sources based on the report receiving rules, and to uniformly parse and structure the collected financial product operation reports to generate a structured product tag containing a unique product identifier. The financial product operation report is then associated with and stored with the structured product tag. The preset sources include at least one of email, internal marketing system, and file upload interface. The acquisition unit is used to determine a set of target users with report viewing permissions and acquire environmental fingerprint information of the terminal used by each target user in the set of target users, wherein the environmental fingerprint information includes at least one of network status, device performance parameters, and user interface focus status. The information push unit is used to decompose the financial product operation report into multiple logical sub-streams based on the content of the associated stored financial product operation report, and generate a personalized streaming transmission strategy for each target user based on the environmental fingerprint information. Based on the streaming transmission strategy, the logical sub-streams are pushed to the corresponding target user's terminal. The logical sub-streams include at least a core data stream with priority and sensitivity, a visualization chart data stream, and a full document file data stream. The streaming transmission strategy defines at least one of the following: transmission order, compression ratio, and start / stop control rules for each logical sub-stream.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images