A Big Data-Based Intelligent Analysis Method for Financial Risk

Abstract

This invention discloses a big data-based intelligent financial risk analysis method, belonging to the field of big data analysis technology. The method includes: refining a standardized risk event set to form financial risk intensity sequences, contract risk intensity sequences, and supply chain risk intensity sequences within a time window; accumulating these sequences within the time window to form financial risk deposition quantity sequences, contract risk deposition quantity sequences, and supply chain risk deposition quantity sequences, and generating linkage trigger signals; and based on these trigger signals, performing reverse backtracking on the financial risk deposition quantity sequences, contract risk deposition quantity sequences, and supply chain risk deposition quantity sequences along a time dimension to identify risk source sequences and diffusion relationships, thus constructing a risk tracing chain. This invention achieves dynamic tracing of the tracing origin and diffusion relationship chain.

Description

Technical Field

[0001] This invention relates to the field of big data analytics, and in particular to a method for intelligent financial risk analysis based on big data. Background Technology

[0002] In business operations, financial status, contract performance, and supply chain operation are intertwined, forming a complex risk transmission network. As business scale expands and transaction stages increase, risk identification and analysis have evolved from single-dimensional financial monitoring to multi-dimensional data analysis encompassing finance, contracts, and the supply chain. Conventional methods typically rely on multi-source data collection and statistical analysis in a big data environment, combined with time series modeling, correlation assessment, and trend analysis, to quantitatively describe and periodically assess corporate risks. Big data-driven financial risk research methods can provide corporate managers with multi-faceted risk information support, enabling corporate risk management to gradually transition from experience-based judgment to data-driven rational decision-making.

[0003] In practical applications, conventional methods often fail to fully reveal the interconnectedness and evolutionary trends of risks across different dimensions: on the one hand, financial, contractual, and supply chain risks are often quantified and tracked separately, lacking a systematic model of their interactions, resulting in an inaccurate grasp of the spread of risks across links; on the other hand, the dynamic changes in the accumulation and progression of risks over time lack effective tracing, making it difficult to identify potential risk sources and spread paths, thus affecting the timeliness and accuracy of overall risk identification. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a big data-based intelligent financial risk analysis method to solve the problems of difficulty in revealing the multi-dimensional linkage and evolution trend of risks, as well as insufficient tracing of risk sources and diffusion paths.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] This invention provides a big data-based intelligent financial risk analysis method, which includes collecting enterprise operation chain data and standardizing the collected enterprise operation chain data to form a standardized risk event set;

[0008] The standardized risk event set is processed to refine the data, forming financial risk intensity sequences, contract risk intensity sequences, and supply chain risk intensity sequences within a time window;

[0009] Within the time window, the financial risk intensity sequence, contract risk intensity sequence, and supply chain risk intensity sequence are accumulated to form the financial risk deposition sequence, contract risk deposition sequence, and supply chain risk deposition sequence, and a linkage trigger signal is generated.

[0010] Based on the linkage trigger signal, the risk accumulation sequence of financial risk, contract risk and supply chain risk is traced back in the time dimension to identify the risk source sequence and diffusion relationship and build a risk tracing chain.

[0011] By analyzing the sequence of risk sources, the number of diffusion relationships, and the accelerating trend in the risk tracing chain, the overall risk level of enterprise operations is classified, and risk warning results are generated.

[0012] As a preferred embodiment of the big data-based intelligent financial risk analysis method described in this invention, the collected enterprise operation chain data includes financial data, contract data, and supply chain data.

[0013] As a preferred embodiment of the big data-based intelligent financial risk analysis method of the present invention, the standardization process to form a standardized risk event set includes unifying the time fields of financial data, contract data, and supply chain data into a standard time format of the same time zone and to the same time precision.

[0014] After unifying the time field, financial risk event sets, contract risk event sets, and supply chain risk event sets are generated separately, and data integrity processing is performed, including missing data completion, duplicate data removal, and anomaly data removal.

[0015] After data integrity processing is completed, the financial risk event set, contract risk event set, and supply chain risk event set are aligned on the timeline and merged to form a standardized risk event set.

[0016] As a preferred embodiment of the big data-based intelligent financial risk analysis method of the present invention, the data refinement processing of the standardized risk event set includes dividing the standardized risk event set into periods within a time window, and extracting financial risk event sets, contract risk event sets, and supply chain risk event sets for each period.

[0017] Calculate the risk indicators corresponding to the financial status for the set of financial risk events, and arrange them in chronological order to form a sequence of financial risk intensity.

[0018] Calculate the risk indicators corresponding to the contract execution status for the set of contract-related risk events, and arrange them in chronological order to form a sequence of contract-related risk intensity.

[0019] Calculate the risk indicators corresponding to the supply chain operation status for the set of supply chain risk events, and arrange them in chronological order to form a supply chain risk intensity sequence;

[0020] In the process of calculating the period of the risk intensity sequence, when any risk indicator is missing, the value of the missing risk indicator in the current period is recorded as zero.

[0021] As a preferred embodiment of the big data-based intelligent financial risk analysis method of the present invention, the step of performing cumulative processing to form a sequence of financial risk deposition, a sequence of contract risk deposition, and a sequence of supply chain risk deposition includes, in each of the various risk intensity sequences, combining risk indicators within the same risk intensity sequence into a risk measurement value by adding them one by one in each period.

[0022] The risk measurement values ​​for each period are accumulated period by period, and the accumulated risk measurement values ​​are arranged in chronological order to form a sequence of financial risk accumulation, a sequence of contractual risk accumulation, and a sequence of supply chain risk accumulation.

[0023] As a preferred embodiment of the big data-based intelligent financial risk analysis method of the present invention, the generation of linkage trigger signals includes selecting, from various risk deposition sequence, a number of consecutive historical periods with the current period as the base point to form a financial deposition history set, a contract deposition history set, and a supply chain deposition history set.

[0024] For each set of sedimentary histories, the median of all cumulative risk measures is calculated as the sedimentary median. Based on the sedimentary median, the deviation of each cumulative risk measure from the sedimentary median is calculated, and the deviation is averaged to obtain the mean deviation.

[0025] The difference between the upper quartile and the median of sedimentary histories for each set of sedimentary histories is calculated to obtain the upper quartile difference.

[0026] Compare the mean deviation with the upper quantile difference and select the larger value as the offset.

[0027] The median value of the sedimentation and the offset are summed to obtain the financial trigger threshold, contract trigger threshold and supply chain trigger threshold respectively;

[0028] The cumulative risk measurement values ​​for each period are compared with the financial trigger threshold, contract trigger threshold, and supply chain trigger threshold respectively to generate financial risk trigger signals, contract risk trigger signals, and supply chain risk trigger signals.

[0029] At the end of each period, financial risk trigger signals, contract risk trigger signals, and supply chain risk trigger signals are compared and linked to generate financial contract linkage trigger signals, financial supply chain linkage trigger signals, contract supply chain linkage trigger signals, and all-dimensional linkage trigger signals.

[0030] As a preferred embodiment of the big data-based intelligent financial risk analysis method of the present invention, the reverse backtracking includes taking the cumulative risk measurement value of the corresponding risk deposition sequence in the corresponding period as the starting point for tracing back when any kind of linkage trigger signal is established.

[0031] After the starting point of the source tracing is determined, the difference between the cumulative risk measurement values ​​of two adjacent periods is calculated along the time dimension, and the difference is defined as the risk increment value.

[0032] When the risk increment value remains positive in consecutive periods, the corresponding risk deposition sequence is marked as the risk source sequence;

[0033] When the absolute value of the risk increment in the risk source sequence gradually increases, it is determined that the risk source sequence has an accelerating increasing trend.

[0034] As a preferred embodiment of the big data-based intelligent financial risk analysis method described in this invention, the construction of the risk tracing chain includes comparing the sequential increasing relationship of different types of risk deposition sequences in the time dimension. When a certain risk deposition sequence increases continuously before another risk deposition sequence, the former risk deposition sequence is marked as the diffusion source, and the latter risk deposition sequence is marked as the diffusion target, thus forming a diffusion relationship.

[0035] Following the chronological order, the diffusion relationship from the diffusion source to the diffusion target is sequentially connected with the risk source sequence, and the risk tracing chain is formed segment by segment until all the risk deposition sequence involved in the linkage trigger signal is included in the risk tracing chain.

[0036] When a risk source sequence is determined to have an accelerating increasing trend, the risk source sequence is marked as an accelerating increasing sequence.

[0037] As a preferred embodiment of the big data-based intelligent financial risk analysis method of the present invention, the classification of the overall risk level of enterprise operations includes determining the overall risk level of enterprise operations as low risk when the risk tracing chain contains only a single category of risk deposition sequence and there is no accelerating increasing trend.

[0038] When the risk tracing chain contains at least one diffusion relationship, or when the risk source sequence shows an accelerating increasing trend, the overall risk level of the enterprise's business activities is determined to be medium risk.

[0039] When the risk tracing chain covers all categories of risk deposition sequences, and at least one risk source sequence shows an accelerating increasing trend, the overall risk level of the enterprise's business activities is determined to be high-level risk.

[0040] As a preferred embodiment of the big data-based intelligent financial risk analysis method described in this invention, the risk warning result includes outputting a risk tracing chain and a risk level.

[0041] The beneficial effects of this invention are as follows: by constructing a dynamic trigger threshold determination mechanism in the time dimension, the comparison and signal triggering of multi-category risk deposition sequence are realized, thereby enabling the identification of linkage between different risk dimensions; by performing reverse backtracking and diffusion relationship tracking under linkage trigger signal conditions, the dynamic tracing of the source starting point and diffusion relationship link is realized. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart of a big data-based intelligent financial risk analysis method.

[0044] Figure 2 A flowchart for generating a linkage trigger signal.

[0045] Figure 3 This is a flowchart for reverse backtracking.

[0046] Figure 4 A flowchart for constructing a risk tracing chain. Detailed Implementation

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0048] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0050] Reference Figures 1-4 This is one embodiment of the present invention, which provides a big data-based intelligent financial risk analysis method, including the following steps:

[0051] S1. Collect enterprise operation chain data and standardize the collected enterprise operation chain data to form a standardized risk event set.

[0052] Furthermore, data collected from the enterprise's operational chain includes financial data, contract data, and supply chain data.

[0053] Financial data includes cash inflow records, cash outflow records, accounts receivable balance, accounts receivable aging distribution, invoicing records, collection records, as well as the amounts of current assets, inventory, and current liabilities in the financial statements, which are used to reflect the company's financial status and the degree of financial pressure.

[0054] Contract data includes contract amount, contract signing date, agreed completion date, actual completion date, and execution status of breach of contract clauses, which are used to reflect the contract execution status.

[0055] Supply chain data includes estimated arrival date, actual arrival date, inventory quantity, inventory inbound time, and inventory outbound time, which are used to reflect the operational status of the supply chain and its supply guarantee capabilities.

[0056] Furthermore, after collecting data on the enterprise's operational chain, the time fields of financial data, contract data, and supply chain data are unified into a standard time format in the same time zone and with the same time precision.

[0057] After the time field is unified, financial risk event sets, contract risk event sets, and supply chain risk event sets are generated respectively.

[0058] After the risk event set is generated, data integrity processing is performed on the financial risk event set, contract risk event set, and supply chain risk event set, including missing data completion, duplicate removal, and anomaly removal. Specifically, missing data records in each risk event set are supplemented with valid values ​​from adjacent time points, only the latest data record is retained for data records that appear repeatedly within a short period of time, and data records that do not conform to the normal value range are removed.

[0059] After data integrity processing is completed, the financial risk event set, contract risk event set, and supply chain risk event set are aligned on the timeline and finally merged to form a standardized risk event set.

[0060] S2. Refine the standardized risk event set to form financial risk intensity sequences, contract risk intensity sequences, and supply chain risk intensity sequences within a time window.

[0061] Furthermore, the process divides the time frame into periods, for example, each month or quarter. Within each period, financial risk events, contractual risk events, and supply chain risk events are extracted from the standardized risk event set. For each type of risk event, data is refined, and risk indicators corresponding to financial status, contract execution status, and supply chain operation status are calculated for each item. After arranging the risk indicators obtained in each period in chronological order, financial risk intensity sequences, contractual risk intensity sequences, and supply chain risk intensity sequences are formed. The specific steps are as follows:

[0062] In cases of concentrated financial risk events, cash inflow and outflow records are compared periodically to calculate the net cash difference. The proportion of this net cash difference to the total cash inflow is used as a cash pressure indicator to reflect whether the company is experiencing excessive cash flow depletion. A higher cash pressure indicator indicates greater cash flow pressure. For accounts receivable balance and aging distribution, the overdue days are used as a benchmark. Accounts receivable exceeding the overdue days are accumulated, and the proportion of this accumulated amount to the total accounts receivable balance is used as an overdue receivable indicator to reflect the risk of delayed collection. A higher overdue receivable indicator indicates a higher risk of delayed collection. For invoicing and payment records, the invoice date and corresponding actual payment date are compared to calculate the payment interval for all invoices. The payment interval is then further analyzed. The following methods are used to calculate the average collection interval in days: First, average collection interval in days is calculated, and the ratio of this average to the company's internally managed collection interval is used as the collection lag indicator. This indicator measures whether there is a delay in the company's cash collection efficiency; a larger collection lag indicator indicates lower cash collection efficiency. Second, the amounts of current assets and current liabilities in the financial statements are compared period by period to calculate the difference between them, resulting in the debt-to-asset ratio. The ratio of this ratio to current liabilities is used as the debt service gap indicator, reflecting the company's ability to repay current liabilities in the short term. A negative debt service gap indicator indicates potential short-term debt repayment pressure. Third, the amounts of inventory and current assets are compared, and the ratio of inventory to current assets is used as the inventory pressure indicator. This indicator reflects the degree to which current assets are tied up in inventory during operations; a larger inventory pressure indicator indicates a higher risk of excessive current assets being tied up in inventory.

[0063] Among them, the values ​​of the funding pressure indicator, overdue receivables indicator, collection lag indicator and inventory pressure indicator are all greater than or equal to zero, while the value of the debt repayment gap indicator can be any real number.

[0064] The financial risk intensity sequence is formed by arranging the funding pressure indicators, overdue receivables indicators, collection lag indicators, debt repayment gap indicators, and inventory pressure indicators obtained in each period in chronological order.

[0065] In the contract-related risk event cluster, the contract signing date, agreed completion date, and actual completion date are compared to calculate the time difference between the actual completion date and the agreed completion date, resulting in the number of days of performance delay. The proportion of the performance delay days to the agreed completion period is used as a delay risk indicator to reflect the degree of delay in contract execution. The higher the delay risk indicator, the higher the risk of untimely contract performance. Based on the contract amount, the proportion of the contract amount in the total contract amount is calculated to obtain the contract amount percentage value. This contract amount percentage value is used as a contract amount impact indicator to reflect the impact of contract delays on the overall business scale. The higher the contract amount impact indicator, the higher the economic impact of the delay risk. Regarding the execution status of breach of contract clauses, a value of 1 is recorded when a breach of contract clause is executed, and a value of 0 is recorded when a breach of contract clause is not executed. The value corresponding to the execution status of breach of contract clauses is used as a breach of contract risk indicator to reflect the liability risk in the contract performance process. A breach of contract risk indicator value of 1 indicates that the company has actually assumed breach of contract liability, and a breach of contract risk indicator value of 0 indicates that the company has not assumed breach of contract liability.

[0066] Among them, the values ​​of the delay risk indicator and the contract amount impact indicator are both greater than or equal to zero, and the value range of the default risk indicator is a discrete value with a value set of {0, 1}.

[0067] The delay risk indicators, contract amount impact indicators, and default risk indicators obtained in each period are arranged in chronological order to form a sequence of contract risk intensity.

[0068] In the context of concentrated supply chain risk events, the estimated arrival date is compared with the actual arrival date to calculate the time difference between the two dates, resulting in the number of days of delayed delivery. This delayed delivery day ratio, representing a percentage of the procurement cycle, serves as the delayed delivery indicator, reflecting the risk of delivery delays in the supply chain. A higher delayed delivery indicator indicates a higher risk of delivery delays. Combining inventory receipt and receipt times, the time difference between inventory receipt and receipt times is calculated to obtain the number of days of inventory retention. This retention day ratio, representing a percentage of the inventory turnover cycle, serves as the inventory backlog indicator, reflecting inventory turnover efficiency. A higher inventory backlog indicator indicates lower inventory liquidity. Regarding inventory quantity, the inventory quantity is compared with the company's internally managed inventory limit. The ratio of the inventory quantity to this limit serves as the excess inventory indicator, reflecting the risk of inventory exceeding a reasonable range. A higher excess inventory indicator indicates a higher risk of material backlog in the company's supply chain.

[0069] Among them, the values ​​of the delayed delivery indicator, the inventory backlog indicator, and the excess inventory indicator are all greater than or equal to zero.

[0070] The indicators of delayed delivery, backlog, and excess inventory obtained in each period are arranged in chronological order to form a sequence of supply chain risk intensity.

[0071] It should be noted that in the calculation of the risk intensity series, each period should include the financial risk intensity series' capital pressure indicator, overdue receivables indicator, collection lag indicator, debt repayment gap indicator, and inventory pressure indicator (or all risk indicators corresponding to the contract risk intensity series and supply chain risk intensity series); when the data source of a certain period is missing the calculation conditions for any risk indicator, the value of the missing risk indicator in the current period will be directly recorded as zero.

[0072] S3. Within the time window, the financial risk intensity sequence, contract risk intensity sequence, and supply chain risk intensity sequence are accumulated to form the financial risk deposition sequence, contract risk deposition sequence, and supply chain risk deposition sequence, and a linkage trigger signal is generated.

[0073] Furthermore, in the financial risk intensity sequence, contract risk intensity sequence, and supply chain risk intensity sequence, the risk indicators within the same risk intensity sequence are combined into a risk measurement value by adding them one by one in each period. Then, the risk measurement values ​​of each period are accumulated in the time dimension. The cumulative risk measurement values ​​obtained by accumulating them period by period are arranged in chronological order to form the financial risk deposition quantity sequence, contract risk deposition quantity sequence, and supply chain risk deposition quantity sequence.

[0074] For example, in the financial risk intensity sequence, the indicators of funding pressure, overdue receivables, collection lag, debt repayment gap, and inventory pressure included in the first period are added together to obtain the risk measurement value for the first period, assuming the risk measurement value for the first period is 2. In the second period, the indicators of funding pressure, overdue receivables, collection lag, debt repayment gap, and inventory pressure are added together to obtain the risk measurement value for the second period, assuming the risk measurement value for the second period is 3. The risk measurement value 2 in the first period and the risk measurement value 3 in the second period are accumulated to obtain the cumulative risk measurement value 5 in the second period. If the risk measurement value is 4 in the third period, then the risk measurement values ​​2 in the first period, 3 in the second period, and 4 in the third period are accumulated to obtain the cumulative risk measurement value 9 in the third period. All cumulative risk measurement values ​​(including the risk measurement value in the first period) are arranged in chronological order to form the financial risk accumulation sequence.

[0075] Furthermore, within the same time window, trigger thresholds are calculated for the financial risk accumulation sequence, contractual risk accumulation sequence, and supply chain risk accumulation sequence, respectively. The specific steps are as follows:

[0076] In the sequence of financial risk deposits, under the same time window, the cumulative risk measurement values ​​of several consecutive historical periods are selected with the current period as the base point to form a set of financial deposit history. The number of consecutive historical periods is defined as the historical sample length. The historical sample length is consistent with the time unit of the period. For example, when the time window is monthly, the historical sample length can be the period of twelve consecutive months.

[0077] Once the historical set of financial deposits is determined, the median of all cumulative risk measures within the set is calculated to obtain the median value of financial deposits. Then, the absolute value of the difference between each cumulative risk measure and the median value is calculated, and all absolute values ​​are averaged to obtain the mean deviation of financial deposits. The difference between the upper quartile of the historical set and the median value is calculated to obtain the upper quartile difference of financial deposits. The mean deviation of financial deposits is compared with the upper quartile difference, and the larger value is selected as the financial offset. Finally, the median value and the offset of financial deposits are summed to obtain the financial trigger threshold.

[0078] In the contract risk deposition sequence and the supply chain risk deposition sequence, the same trigger threshold calculation method as the financial risk deposition sequence is used to obtain the contract trigger threshold and the supply chain trigger threshold, respectively.

[0079] It should be noted that when the historical sample length is insufficient, the same calculation process is performed using the cumulative risk measure values ​​of all currently available periods; when there is only one period, the risk measure value of that period is compared with zero, and the larger value is selected as the trigger threshold.

[0080] It should also be noted that, as the financial, contractual, and supply chain historical data sets are continuously updated over time windows, the trigger thresholds for these categories will also be dynamically adjusted accordingly.

[0081] Furthermore, the cumulative risk measurement values ​​for corresponding periods in the financial risk accumulation series, contract risk accumulation series, and supply chain risk accumulation series are compared with the financial trigger threshold, contract trigger threshold, and supply chain trigger threshold, respectively, to generate financial risk trigger signals, contract risk trigger signals, and supply chain risk trigger signals, respectively. The specific steps are as follows:

[0082] In each period, when the cumulative risk measure value of the financial risk accumulation sequence is greater than the financial trigger threshold, a financial risk trigger signal is recorded; when the cumulative risk measure value of the contract risk accumulation sequence is greater than the contract trigger threshold, a contract risk trigger signal is recorded; when the cumulative risk measure value of the supply chain risk accumulation sequence is greater than the supply chain trigger threshold, a supply chain risk trigger signal is recorded.

[0083] At the end of each period, financial risk trigger signals, contract risk trigger signals, and supply chain risk trigger signals are stored as the trigger results for this period and arranged in chronological order to form a trigger signal sequence; as the time window continues to advance, the trigger results for each period will be dynamically updated.

[0084] Furthermore, financial risk trigger signals, contractual risk trigger signals, and supply chain risk trigger signals are linked and compared to form financial-contract linkage trigger signals, financial-supply chain linkage trigger signals, contract-supply chain linkage trigger signals, and all-dimensional linkage trigger signals, respectively. The specific steps are as follows:

[0085] In each period, when both financial and contractual risk trigger signals are triggered simultaneously, a financial-contractual linkage trigger signal is recorded. This indicates that the company is under financial pressure while simultaneously facing contract delay risks, suggesting the possibility of a combination of a broken cash flow and contract failure. When both financial and supply chain risk trigger signals are triggered simultaneously, a financial-supply chain linkage trigger signal is recorded. This indicates that the company is under financial pressure while simultaneously facing supply chain delays or inventory backlog risks, suggesting the possibility of insufficient funds leading to supply chain disruptions. When both contractual and supply chain risk trigger signals are triggered simultaneously, a contract-supply chain linkage trigger signal is recorded. This indicates the interaction between contract performance delays and supply chain delivery delays, suggesting the possibility of performance obstruction. When all three risk trigger signals (financial, contractual, and supply chain) are triggered simultaneously, a full-dimensional linkage trigger signal is recorded, indicating a severe risk state where the company is simultaneously unbalanced across the three dimensions of capital, contracts, and supply chain.

[0086] At the end of each period, the financial contract linkage trigger signal, the financial supply chain linkage trigger signal, the contract supply chain linkage trigger signal, and the all-dimensional linkage trigger signal are stored together as the comprehensive trigger result for this period, and arranged in chronological order to form a comprehensive trigger signal sequence; as the time window continues to advance, the comprehensive trigger result is dynamically updated with each period.

[0087] S4. Based on the linkage trigger signal, the risk accumulation sequence of financial risk, contract risk, and supply chain risk is traced back in the time dimension to identify the risk source sequence and diffusion relationship, and to build a risk traceability chain.

[0088] Furthermore, in each period, when the financial contract linkage trigger signal is established, the cumulative risk measurement values ​​of the corresponding periods in the financial risk deposition sequence and the contract risk deposition sequence are used as the starting point for tracing the source; when the financial supply chain linkage trigger signal is established, the cumulative risk measurement values ​​of the corresponding periods in the financial risk deposition sequence and the supply chain risk deposition sequence are used as the starting point for tracing the source; when the contract supply chain linkage trigger signal is established, the cumulative risk measurement values ​​of the corresponding periods in the contract risk deposition sequence and the supply chain risk deposition sequence are used as the starting point for tracing the source; when the all-dimensional linkage trigger signal is established, the cumulative risk measurement values ​​of the corresponding periods in the financial risk deposition sequence, the contract risk deposition sequence, and the supply chain risk deposition sequence are all used as the joint starting point for tracing the source.

[0089] Furthermore, after the starting point for tracing is determined, the evolution of cumulative risk measurement values ​​for the financial risk accumulation sequence, contract risk accumulation sequence, and supply chain risk accumulation sequence is traced backward along the time dimension. The specific steps are as follows:

[0090] The difference between the cumulative risk measurement values ​​of two adjacent periods is calculated sequentially, and the difference between the cumulative risk measurement values ​​of two adjacent periods is defined as the risk increment value.

[0091] When the risk increment is positive, it means that the cumulative risk measurement value in the next period is higher than that in the previous period, and the risk is increasing; when the risk increment is negative, it means that the cumulative risk measurement value in the next period is lower than that in the previous period, and the risk is decreasing; when the risk increment is zero, it means that the cumulative risk measurement value in the next period is equal to that in the previous period, and the risk remains unchanged.

[0092] When the risk increment value of the financial risk deposition sequence, contract risk deposition sequence, or supply chain risk deposition sequence is greater than zero in multiple consecutive periods, it indicates that the financial risk deposition sequence, contract risk deposition sequence, or supply chain risk deposition sequence has a continuous increasing trend in the time dimension, and the corresponding risk deposition sequence is marked as the risk source sequence.

[0093] It should be noted that when the risk increment value is not only greater than zero but also the absolute value of the risk increment value gradually increases, it indicates that the financial risk deposition sequence, contract risk deposition sequence, or supply chain risk deposition sequence has an accelerating increasing trend in the time dimension.

[0094] After the risk source sequence is identified, the sequential increase relationship of different types of risk deposition sequence in the time dimension is further compared. When there is a certain type of risk deposition sequence that increases continuously before another type of risk deposition sequence in the financial, contractual, and supply chain risk deposition sequence, the risk deposition sequence that appears first and continues to increase is marked as the diffusion source, and the risk deposition sequence that appears later and continues to increase is marked as the diffusion target.

[0095] For example, when the sequence of financial risk deposits increases continuously before the sequence of contractual risk deposits, the sequence of financial risk deposits is marked as the diffusion source, and the sequence of contractual risk deposits is marked as the diffusion target.

[0096] Furthermore, after the risk source sequence and diffusion relationship are determined, the diffusion relationship from the diffusion source to the diffusion target is sequentially connected to the risk source sequence in chronological order, and the risk tracing chain is formed segment by segment. The risk tracing chain starts with the risk source sequence and adds diffusion relationships consisting of diffusion sources and diffusion targets in sequence until all risk deposition sequences involved in all linkage trigger signals are included in the risk tracing chain.

[0097] For example, in consecutive periods, if the sequence of financial risk deposits is identified as the source of risk and continues to increase in subsequent periods; at the beginning of a later period, the sequence of contractual risk deposits shows a continuous increasing trend, so the sequence of financial risk deposits is marked as the diffusion source and the sequence of contractual risk deposits is marked as the diffusion target, forming a diffusion relationship of "financial risk deposit sequence → contractual risk deposit sequence"; in subsequent periods, the sequence of supply chain risk deposits begins to increase continuously, at which point the sequence of contractual risk deposits is marked as the diffusion source and the sequence of supply chain risk deposits is marked as the diffusion target, forming a diffusion relationship of "contractual risk deposit sequence → supply chain risk deposit sequence"; finally, the risk tracing chain generated over the entire time period is: "financial risk deposit sequence → contractual risk deposit sequence → supply chain risk deposit sequence", representing the complete process of risk gradually spreading from financial pressure to contract performance delays, and then further to supply chain delivery obstruction.

[0098] It should be noted that, for cases where multiple diffusion relationships exist within the same period, multiple risk tracing chains are generated and stored separately according to the chronological order and the order in which the diffusion relationships appear. At the same time, for each diffusion relationship, the diffusion coverage range, diffusion source category, diffusion target category, diffusion order, and the range of risk increment values ​​within the periods covered by the diffusion relationship (e.g., the risk increment values ​​between the start and end periods) are recorded, and the risk source sequence is recorded at the risk tracing chain level.

[0099] It should also be noted that during the construction of the risk tracing chain, when the risk source sequence shows an accelerating increasing trend, the risk source sequence will be marked as an accelerating increasing sequence. The accelerating increasing sequence has a higher risk evolution intensity in the risk tracing chain, and its position in the risk tracing chain will be marked in a key way to highlight the accelerated characteristics of the risk diffusion trend.

[0100] S5. By analyzing the sequence of risk sources, the number of diffusion relationships, and the accelerating trend in the risk tracing chain, the overall risk level of the enterprise's business activities is classified, and risk warning results are generated.

[0101] Furthermore, based on the risk source sequence, the number of diffusion relationships, and the marking of accelerating sequences contained in the risk tracing chain, the overall risk level of the enterprise's business activities is classified:

[0102] When the risk tracing chain contains only a single category of risk deposition sequence, and the corresponding risk deposition sequence is not marked as an accelerating increasing sequence, the overall risk level of the enterprise's business activities is determined to be low-level risk, indicating that the risk is still limited to a single business dimension, such as only financial pressure or only contract delays.

[0103] When the risk tracing chain contains at least one diffusion relationship from the diffusion source to the diffusion target, or when the risk source sequence is marked as an accelerating sequence, the overall risk level of the enterprise's business activities is determined to be medium risk, indicating that the risk has spread between different categories, or the risk has accumulated rapidly within a certain category, such as financial pressure spreading to contract delays, or the sequence of financial risk accumulation showing an accelerating increase.

[0104] When the risk tracing chain covers the sequence of financial risk accumulation, contract risk accumulation, and supply chain risk accumulation, and at least one of the risk source sequences is marked as an accelerating increasing sequence, the overall risk level of the enterprise's business activities is determined to be high-level risk. This indicates that the risk has spread comprehensively across the three dimensions of the enterprise's operations: capital, contracts, and supply chain, and the risk evolution trend is accelerating, indicating that the enterprise's operations are in a state of serious imbalance.

[0105] Furthermore, after classifying the overall risk level of a company's business operations, risk warning results are generated.

[0106] The risk warning results include the risk tracing chain and the risk level. The risk tracing chain records the risk source sequence, the spread coverage area, the spread source category, the spread target category, the spread order, and the range of risk increment values ​​within the spread relationship coverage period. The risk level includes low risk, medium risk, and high risk.

[0107] When the linkage trigger signal belongs to the financial contract linkage trigger signal, financial supply chain linkage trigger signal, contract supply chain linkage trigger signal or all-dimensional linkage trigger signal, it should be clearly marked in the risk warning result so as to intuitively present the transmission relationship between the financial risk accumulation sequence, the contract risk accumulation sequence and the supply chain risk accumulation sequence.

[0108] Risk warning results are generated and stored periodically and updated in each period as the time window progresses. Historical risk warning results are retained intact for longitudinal comparison of changes in the risk tracing chain, risk level, and risk increment range.

[0109] In summary, this invention achieves the comparison and signal triggering of multi-category risk deposition sequences by constructing a dynamic trigger threshold determination mechanism in the time dimension, thereby enabling the identification of linkage between different risk dimensions; and achieves dynamic tracing of the source starting point and diffusion relationship link by performing reverse backtracking and diffusion relationship tracking under linkage trigger signal conditions.

[0110] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A big data-based financial risk intelligent analysis method, characterized in that: Comprising, Collect enterprise operation link data, and standardize the collected enterprise operation link data to form a standardized risk event set; Refine the data of the standardized risk event set to form a financial risk intensity sequence, a contract risk intensity sequence, and a supply chain risk intensity sequence under a time window; Under the time window, accumulate the financial risk intensity sequence, the contract risk intensity sequence, and the supply chain risk intensity sequence to form a financial risk deposition sequence, a contract risk deposition sequence, and a supply chain risk deposition sequence, and generate a linkage trigger signal; The generation of the linkage trigger signal includes, in each risk deposition sequence, selecting the cumulative risk measurement values of a number of consecutive historical periods with the current period as the base point to form a financial deposition history set, a contract deposition history set, and a supply chain deposition history set; For each deposition history set, calculate the median of all cumulative risk measurement values as a deposition median value, and calculate the deviation of each cumulative risk measurement value from the deposition median value, and average the deviation to obtain a deviation mean value; Calculate the difference between the upper quartile of each deposition history set and the deposition median value to obtain an upper quartile difference; Compare the deviation mean value and the upper quartile difference, and select the larger value as the offset; Sum the deposition median value and the offset to obtain a financial trigger threshold, a contract trigger threshold, and a supply chain trigger threshold, respectively; Compare each period's cumulative risk measurement value with the financial trigger threshold, the contract trigger threshold, and the supply chain trigger threshold to generate a financial risk trigger signal, a contract risk trigger signal, and a supply chain risk trigger signal; At the end of each period, compare the financial risk trigger signal, the contract risk trigger signal, and the supply chain risk trigger signal to generate a financial contract linkage trigger signal, a financial supply chain linkage trigger signal, a contract supply chain linkage trigger signal, and a full-dimensional linkage trigger signal; Based on the linkage trigger signal, reverse trace the financial risk deposition sequence, the contract risk deposition sequence, and the supply chain risk deposition sequence in the time dimension to identify a risk source sequence and a diffusion relationship, and construct a risk traceability chain; Classify the overall risk level of the enterprise's operation activities through the risk source sequence, the number of diffusion relationships, and the accelerating increasing trend in the risk traceability chain, and generate a risk warning result.

2. The big data based financial risk intelligent analysis method of claim 1, wherein: The collection of enterprise operation link data includes financial data, contract data, and supply chain data.

3. The big data based financial risk intelligent analysis method of claim 2, wherein: The standardization processing includes unifying the time fields of the financial data, the contract data, and the supply chain data into the same time zone standard time format and the same time precision; After unifying the time fields, generate a financial risk event set, a contract risk event set, and a supply chain risk event set, and perform data integrity processing of missing completion, duplicate removal, and abnormal removal; After the data integrity processing is completed, the financial risk event set, the contract risk event set and the supply chain risk event set are aligned on the time axis and merged to form a standardized risk event set.

4. The big data based financial risk intelligent analysis method of claim 3, wherein: The data refinement processing on the standardized risk event set comprises dividing the standardized risk event set into periods under a time window, and extracting the financial risk event set, the contract risk event set and the supply chain risk event set respectively in each period; The risk indicators corresponding to the fund status of the financial risk event set are calculated and arranged in time sequence to form a financial risk intensity sequence; The risk indicators corresponding to the contract execution status of the contract risk event set are calculated and arranged in time sequence to form a contract risk intensity sequence; The risk indicators corresponding to the supply chain operation status of the supply chain risk event set are calculated and arranged in time sequence to form a supply chain risk intensity sequence; In the period calculation process of the risk intensity sequence, when any risk indicator is missing, the value of the missing risk indicator in the current period is recorded as zero.

5. The big data based financial risk intelligent analysis method of claim 4, wherein: The cumulative processing comprises adding the risk indicators in each period in the same risk intensity sequence to form a risk measurement value; The risk measurement values in each period are accumulated, and the accumulated risk measurement values obtained by the period-by-period accumulation are arranged in time sequence to form the financial risk deposition sequence, the contract risk deposition sequence and the supply chain risk deposition sequence.

6. The big data based financial risk intelligent analysis method of claim 5, wherein: The reverse tracing comprises, when any kind of linkage trigger signal is established, taking the accumulated risk measurement value of the corresponding risk deposition sequence in the corresponding period as the tracing starting point; After the tracing starting point is determined, the difference between the accumulated risk measurement values of two adjacent periods is calculated period by period along the time dimension, and the difference is defined as a risk increment value; When the risk increment value remains positive in consecutive periods, the corresponding risk deposition sequence is marked as a risk source sequence; When the absolute value of the risk increment value of the risk source sequence gradually increases, it is determined that the risk source sequence has an accelerating increasing trend.

7. The big data based financial risk intelligent analysis method of claim 6, wherein: The construction of the risk tracing chain comprises comparing the increasing relationship of different categories of risk deposition sequences in the time dimension, when a certain risk deposition sequence continuously increases before another risk deposition sequence, marking the former risk deposition sequence as a diffusion source and the latter risk deposition sequence as a diffusion target, and forming a diffusion relationship; The diffusion relationship from the diffusion source to the diffusion target is sequentially connected with the risk source sequence in time sequence to form the risk tracing chain in segments, until all the risk deposition sequences involved in the linkage trigger signal are included in the risk tracing chain; When it is determined that the risk source sequence has an accelerating increasing trend, the risk source sequence is marked as an accelerating increasing sequence.

8. The big data based financial risk intelligent analysis method of claim 7, wherein: The classification of the overall risk level of the enterprise's business activities comprises, when the risk tracing chain only contains a single category of risk deposition sequence and no accelerating increasing trend appears, determining that the overall risk level of the enterprise's business activities is low risk. When the risk traceability chain contains at least one diffusion relationship, or the risk source sequence has an accelerating increasing trend, the overall risk level of the business operation activities is determined as a medium risk level; When the risk traceability chain covers all categories of risk deposit amount sequences, and at least one risk source sequence has an accelerating increasing trend, the overall risk level of the business operation activities is determined as a high risk level.

9. The big data based financial risk intelligent analysis method of claim 8, wherein: The risk early warning result includes outputting the risk traceability chain and the risk level.

Images