Method and system for identifying risks of marine cargo based on dynamic weight and factor correction

Abstract

The application discloses a kind of sea freight risk identification method and system based on dynamic weight and factor correction, it is related to risk detection technical field.Therein, the method includes: obtaining the risk assessment attribute of target sea freight in at least one risk assessment dimension, and determine the weight to be used corresponding to each risk assessment dimension;Based on the risk assessment attribute and the weight to be used associated with each risk assessment dimension, determine the risk assessment attribute to be adjusted corresponding to target sea freight;Based on the risk factor corresponding to at least one risk influencing factor associated with target sea freight, obtain target correction coefficient;Based on the risk assessment attribute to be adjusted and target correction coefficient, determine the cargo risk level corresponding to target sea freight.Solve the problem that it is difficult to accurately determine the cargo risk level of cargo in the prior art, by combining multiple risk assessment dimensions and risk influencing factors, the effect of accurately determining the cargo risk level of target sea freight is realized.

Description

Technical Field

[0001] This invention relates to the field of risk detection technology, and in particular to a method and system for identifying risks in maritime cargo based on dynamic weights and factor correction. Background Technology

[0002] In maritime transport practice, cargoes are diverse and often possess sensitive or risky attributes, making accurate assessment of cargo risk levels crucial. However, existing cargo risk assessment mechanisms have significant limitations when integrating multi-source data: on the one hand, they lack a unified assessment benchmark, and on the other hand, they cannot flexibly adapt to diverse risk influencing factors, resulting in inaccurate identification of cargo risks.

[0003] To address the aforementioned issues, improvements are needed in the methods for identifying risks associated with maritime cargo. Summary of the Invention

[0004] This invention provides a method for identifying maritime cargo risks based on dynamic weights and factor corrections, in order to solve the problem of difficulty in accurately determining the risk level of cargo.

[0005] In a first aspect, embodiments of the present invention provide a method for identifying maritime cargo risks based on dynamic weights and factor corrections, including: Obtain the risk assessment attributes of the target seaborne cargo under at least one risk assessment dimension, and determine the weights to be used for each risk assessment dimension; Based on the risk assessment attributes and weights to be used associated with each of the aforementioned risk assessment dimensions, the risk assessment attributes to be adjusted for the target seaborne cargo are determined. Based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo, the target correction coefficient is obtained; Based on the risk assessment attributes to be adjusted and the target correction coefficient, the cargo risk level corresponding to the target seaborne cargo is determined.

[0006] Secondly, embodiments of the present invention also provide a maritime cargo risk identification system based on dynamic weights and factor correction, comprising: The weight determination module is used to obtain the risk assessment attributes of the target seaborne cargo under at least one risk assessment dimension, and determine the weights to be used for each risk assessment dimension. The risk assessment attribute determination module is used to determine the risk assessment attribute to be adjusted for the target seaborne cargo based on the risk assessment attributes associated with each of the risk assessment dimensions and the weights to be used. The target correction coefficient determination module is used to obtain the target correction coefficient based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo; The cargo risk level determination module is used to determine the cargo risk level corresponding to the target seaborne cargo based on the risk assessment attribute to be adjusted and the target correction coefficient.

[0007] The technical solution of this invention involves obtaining the risk assessment attributes of a target maritime cargo under at least one risk assessment dimension and determining the weights to be used for each risk assessment dimension; determining the risk assessment attributes to be adjusted for the target maritime cargo based on the risk assessment attributes associated with each risk assessment dimension and the weights to be used; obtaining a target correction coefficient based on the risk factors corresponding to at least one risk influencing factor associated with the target maritime cargo; and determining the cargo risk level of the target maritime cargo based on the risk assessment attributes to be adjusted and the target correction coefficient. This technical solution first determines the risk assessment attributes to be adjusted for the target maritime cargo under multiple risk assessment dimensions, then combines the risk detection conditions corresponding to the risk influencing factors that may be triggered during the actual transportation process of the target maritime cargo, and calculates the target correction coefficient based on the risk factors corresponding to the triggered risk influencing factors. Based on this, the target correction coefficient is used to correct the parameters of the risk assessment attributes to be adjusted to obtain the target risk assessment attributes corresponding to the target maritime cargo, and then the cargo risk level of the target maritime cargo is determined based on the risk assessment attribute range to which the target risk assessment attributes belong. This solves the problem in the prior art of accurately determining the cargo risk level of cargo, and achieves the effect of accurately determining the cargo risk level of the target maritime cargo by combining multiple risk assessment dimensions and risk influencing factors. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the content of the embodiments of the present invention and these drawings without creative effort.

[0009] Figure 1 This is a flowchart of a method for identifying maritime cargo risks based on dynamic weights and factor correction, provided in Embodiment 1 of the present invention. Figure 2 This is a flowchart of a maritime cargo risk identification method based on dynamic weights and factor correction provided in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of a maritime cargo risk identification system based on dynamic weights and factor correction, provided in Embodiment 3 of the present invention. Detailed Implementation

[0010] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention. The acquisition, transmission, storage, use, and processing of data in the technical solutions of this application comply with the relevant provisions of national laws and regulations. It should be noted that in the embodiments of this application, certain software, components, or models and other existing solutions in the industry may be mentioned. These should be considered as exemplary, and their purpose is only to illustrate the feasibility of implementing the technical solutions of this application, but it does not mean that the applicant has or necessarily used such solutions.

[0011] It should be noted that the collection, gathering, updating, analysis, processing, use, transmission, and storage of user personal information involved in the technical solution disclosed herein all comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken to prevent unauthorized access to user personal information data and to maintain user personal information security and network security. It should also be noted that the collection, gathering, updating, analysis, processing, use, transmission, and storage of user personal information involved in the technical solution disclosed herein are all conducted with the user's knowledge and consent, and comply with relevant privacy protection regulations.

[0012] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in sequences other than those illustrated or described herein.

[0013] Before elaborating on this technical solution in detail, a brief introduction to its application scenarios will be provided to facilitate a clearer understanding. This solution can be applied to scenarios where risk level assessments are conducted on goods before transportation. Optionally, this solution can be applied to scenarios where risk level assessments are conducted on ocean freight.

[0014] Example 1 Figure 1The present invention provides a flowchart of a method for identifying maritime cargo risks based on dynamic weights and factor corrections in Embodiment 1. This embodiment is applicable to scenarios where risk level assessments are performed on transported cargo before transportation. The method can be executed by a maritime cargo risk identification system based on dynamic weights and factor corrections. This maritime cargo risk identification system based on dynamic weights and factor corrections can be implemented in hardware and / or software. This maritime cargo risk identification system based on dynamic weights and factor corrections can be configured in a computing device capable of executing the maritime cargo risk identification method based on dynamic weights and factor corrections.

[0015] like Figure 1 As shown, the method includes: S110. Obtain the risk assessment attributes of the target seaborne cargo under at least one risk assessment dimension, and determine the weights to be used for each risk assessment dimension.

[0016] In this context, the target ocean freight can be understood as the primary object requiring risk level assessment. The target ocean freight can be any type of cargo or cargo in any condition, and it is the core assessment object around which the entire risk assessment process revolves. By conducting a risk level assessment on the target ocean freight, it is determined whether any risks exist during the transportation process and, if so, the level of those risks.

[0017] Risk assessment dimensions refer to the factors used to measure, analyze, or judge the characteristics and degree of risk when assessing the risk level of a target maritime cargo. For example, risk assessment dimensions could include the cargo category of the target maritime cargo, cargo hold information carrying the target maritime cargo, historical booking bills of lading information, customer profile dimensions of the users corresponding to the target maritime cargo, and spatiotemporal risk analysis information associated with the target maritime cargo. In practical applications, risk assessment dimensions can be customized according to actual needs; this technical solution does not limit the specific dimensional information of the risk assessment dimensions.

[0018] Risk assessment attributes can be understood as specific quantitative values ​​used to measure the risk level of a target maritime cargo under different risk assessment dimensions. The weights to be used can characterize the degree of influence of each risk assessment dimension on the assessment result of the cargo risk level of the target maritime cargo. A larger weight indicates a greater influence of the corresponding risk assessment dimension on the assessment result of the cargo risk level, and vice versa.

[0019] Specifically, to accurately assess the risk level of target maritime cargo, this technical solution pre-constructs an assessment model containing multiple risk assessment dimensions and their corresponding weights. During actual assessment, the risk assessment attributes of the target maritime cargo under each risk assessment dimension are first obtained, and the weights corresponding to each risk assessment dimension are determined. Then, a correlation calculation (such as weighted summation) is performed based on the risk assessment attributes of each risk assessment dimension and the corresponding weights to comprehensively quantify and determine the risk level of the target maritime cargo.

[0020] S120. Based on the risk assessment attributes and weights to be used associated with each risk assessment dimension, determine the risk assessment attributes to be adjusted for the target seaborne cargo.

[0021] Among them, the risk assessment attributes to be adjusted can be understood as intermediate values ​​initially determined based on the risk assessment attributes and the weights to be used.

[0022] It should be noted that this scheme first calculates the risk assessment attributes to be adjusted for the target seaborne cargo by using the risk assessment attributes and weights associated with each risk assessment dimension. However, these risk assessment attributes to be adjusted are only preliminary identification results and cannot be directly used as the basis for determining the cargo risk level. The final cargo risk level can only be determined after adjustments based on external factors.

[0023] The reason for this is that the cargo risk assessment in this technical solution needs to take into account two layers of influence: firstly, the inherent risks brought about by each risk assessment dimension itself to the risk assessment of the target seaborne cargo; and secondly, the dynamic risk influencing factors existing throughout the entire cargo transportation process. Based on this, after initially generating the risk assessment attributes to be adjusted, this solution needs to correct the risk assessment attributes to be adjusted in combination with risk influencing factors in order to comprehensively determine the cargo risk level of the target seaborne cargo.

[0024] Optionally, based on the risk assessment attributes and weights to be used associated with each risk assessment dimension, the risk assessment attributes to be adjusted for the target seaborne cargo are determined, including: for each risk assessment dimension, a first attribute is obtained based on the risk assessment attributes and weights to be used associated with the current risk assessment dimension; the first attributes corresponding to all risk assessment dimensions are superimposed to obtain the risk assessment attributes to be adjusted for the target seaborne cargo.

[0025] The first attribute is used to characterize the quantitative assessment value of the target maritime cargo under the current risk assessment dimension. The first attribute can be obtained by multiplying the risk assessment attribute of the current risk assessment dimension with the weight to be used.

[0026] For example, for the five risk assessment dimensions associated with the target maritime cargo, the risk assessment attributes and weights to be used for each risk assessment dimension are obtained. Taking the first risk assessment dimension as the current risk assessment dimension, the first attribute is obtained by multiplying the risk assessment attribute and weight to be used for the current risk assessment dimension. Similarly, the first attributes for each of the five risk assessment dimensions are calculated, and all the first attributes are superimposed to obtain the risk assessment attribute to be adjusted for the target maritime cargo.

[0027] S130. Based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo, obtain the target correction coefficient.

[0028] Among them, risk influencing factors can be used to characterize the risk triggers that may be engaged in assessing the risk level of target maritime cargo. Risk factors characterize the degree of influence of risk influencing factors on the assessment of cargo risk level. A larger risk factor indicates a greater impact of the risk influencing factor on the assessment result of the cargo risk level, while a smaller risk factor indicates a smaller impact. The target correction coefficient is used to characterize the adjustment parameter used to correct the risk assessment attribute to be adjusted.

[0029] Optionally, a target correction coefficient is obtained based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo, including: if the target seaborne cargo triggers the risk detection condition corresponding to a single risk influencing factor, then the risk factor corresponding to the risk influencing factor is used as the target correction coefficient.

[0030] In practical applications, the target seaborne cargo may actually trigger one or more risk factors, each of which corresponds to a pre-set risk factor.

[0031] When a target seaborne cargo triggers a risk detection condition corresponding to a single risk influencing factor, the risk factor corresponding to that risk influencing factor is directly used as the target correction coefficient. For example, if the risk factor corresponding to that risk influencing factor is 1.1, then the target correction coefficient is 1.1.

[0032] It should be noted that, in order to strictly adhere to the upper limit of risk factors when the risk detection conditions of a single risk factor are triggered for the target seaborne cargo, the risk factor corresponding to each risk factor can be set to be the same as the upper limit of risk factors.

[0033] Optionally, a target correction coefficient is obtained based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo, including: if the target seaborne cargo triggers risk detection conditions corresponding to at least two risk influencing factors, then a first factor is determined based on the product of the risk factors of all risk influencing factors; an upper limit risk factor is dynamically determined based on the factor correlation between at least two risk influencing factors; if the first factor is less than or equal to the upper limit risk factor, then the first factor is used as the target correction coefficient; if the first factor is greater than the upper limit risk factor, then the upper limit risk factor is used as the target correction coefficient.

[0034] For example, taking five potential risk factors affecting the target ocean cargo as an example: First, the route risk factor corresponding to the target ocean cargo, the risk detection condition for which is: whether any incidents of false or concealed declarations have occurred on the route through which the target ocean cargo travels in the past 30 days; Second, the sensitivity factor of the cargo type corresponding to the target ocean cargo, the risk detection condition for which is: whether the target ocean cargo is a cargo type with high recent frequency of violations; Third, the customer behavior risk factor associated with the target ocean cargo, the risk detection condition for which is: whether the customer activity assessment attribute associated with the target ocean cargo has improved compared to the previous period (wherein, the customer activity assessment attribute can be determined based on the consumption frequency corresponding to the customer profile, or based on information such as historical cargo orders); Fourth, the seasonality risk factor affecting the target ocean cargo, the risk detection condition for which is: whether the cargo type to which the target ocean cargo belongs is in a historically high-incidence period of violations in the current quarter; Fifth, the multi-dimensional resonance risk factor, the risk detection condition for which is: the target ocean cargo involves multiple risk factors simultaneously, such as involving three risk factors simultaneously.

[0035] Based on this, the target maritime cargo is inspected according to the risk detection conditions of various risk influencing factors. For example, if the target maritime cargo has been found to have been falsely reported or concealed within the past 30 days, it is determined that the target maritime cargo has triggered the risk detection condition of the first risk influencing factor. At this time, the risk factor corresponding to the risk influencing factor can be set to 1.2. Conversely, if it has not been triggered, the risk factor corresponding to the risk influencing factor is set to 1.

[0036] For example, if the target seaborne cargo is a type of cargo that has recently been subject to frequent violations, the corresponding risk factor is set at 1.15; conversely, if the target seaborne cargo is not a type of cargo that has recently been subject to frequent violations, the corresponding risk factor is set at 1.

[0037] For example, if the customer activity level of the target ocean freight customer increases by more than 30 points compared to the previous period, the corresponding risk factor is set to 1.1; conversely, if the customer activity level of the target ocean freight customer does not increase by more than 30 points compared to the previous period, the corresponding risk factor is set to 1.

[0038] For example, if the type of cargo to which the target sea freight belongs is in a period of historically high incidence of violations in the current quarter, the corresponding risk factor is determined to be 1.1; conversely, if the type of cargo to which the target sea freight belongs is not in a period of historically high incidence of violations in the current quarter, the corresponding risk factor is determined to be 1.

[0039] In a specific example, if the target seaborne cargo involves risk detection conditions corresponding to two or more risk influencing factors, the first factor is obtained based on the product of the risk factors corresponding to the triggered risk influencing factors. Simultaneously, an upper limit risk factor is determined based on the correlation between the various risk influencing factors, and the target correction coefficient is determined based on the relationship between the first factor and the upper limit risk factor.

[0040] Optionally, the upper limit risk factor is dynamically determined based on the factor correlation between at least two risk influencing factors, including: determining the correlation coefficient between each pair of risk influencing factors and determining the value to be used based on all correlation coefficients; determining the number of effective factors based on the value to be used and the first preset function; and dynamically determining the upper limit risk factor based on the number of effective factors and the second preset function.

[0041] The correlation coefficient is used to characterize the correlation between any two risk factors. A larger correlation coefficient indicates a stronger correlation between the two risk factors, and vice versa. The number of effective factors characterizes the number of identified risk factors that are correlated.

[0042] Specifically, when a target seaborne cargo simultaneously triggers risk detection conditions corresponding to at least two risk influencing factors, in order to avoid the overlapping of risk factors due to the correlation between the various risk influencing factors, the correlation of each risk influencing factor is first calculated. After calculating the correlation coefficients corresponding to each pair of risk influencing factors, the value to be used is obtained based on the average of the absolute values ​​of all correlation coefficients, and the value to be used is denoted as C.

[0043] Based on this, taking the first preset function as K=5 / (1+4×C) and the second preset function as M=1.0+0.2×K as an example, the first preset function is used to obtain the number of effective factors, and the second preset function is used to obtain the upper limit risk factor. The number of effective factors K can be obtained based on the value to be used and the first preset function. For example, when C=0, K=5, indicating that the five risk influencing factors are independent of each other and can be regarded as five different risk signals; when C=1, K=1, indicating that the five risk influencing factors highly overlap, actually reflecting only one risk signal; when C=0.3, K≈2.3, indicating that at least two risk influencing factors are correlated; furthermore, based on the second preset function and the number of effective factors K, the upper limit risk factor M can be obtained.

[0044] The advantage of this setup is that the lower the correlation between risk influencing factors, the more independent the risk signals captured from different dimensions, resulting in a higher upper limit for the product and allowing for the full superposition of multidimensional risks. Conversely, the higher the correlation between risk influencing factors, the more likely they are reflecting different aspects of the same risk, automatically lowering the upper limit for the product and preventing redundant amplification. By introducing a mechanism to dynamically determine the upper limit risk factor, updating it every period (e.g., every quarter) ensures that the upper limit risk factor adapts adaptively to changes in the external risk landscape. Specifically, when adjusting parameters of the risk assessment attributes to be adjusted, if the use of the upper limit risk factor is involved, the updated upper limit risk factor is used as the target adjustment coefficient to adjust the parameters of the risk assessment attributes to be adjusted.

[0045] In this technical solution, the correlation coefficient between any two risk influencing factors can be determined by a correlation calculation algorithm, or it can be obtained by performing correlation analysis on each risk influencing factor based on a correlation analysis model. This technical solution does not limit the method for determining the correlation coefficient. The first preset function and the second preset function can be user-defined functions used to calculate the upper limit risk factor. In practical applications, the first preset function and the second preset function can be set according to actual needs.

[0046] S140. Based on the risk assessment attributes to be adjusted and the target correction coefficient, determine the cargo risk level corresponding to the target seaborne cargo.

[0047] Among them, cargo risk level refers to a quantitative assessment indicator that comprehensively represents the inherent attributes of the target sea freight and the dynamic environment of the transportation process throughout the entire life cycle of logistics and transportation, and assesses the degree of risk of the target sea freight.

[0048] Specifically, determining the cargo risk level corresponding to the target seaborne cargo includes: determining the target risk assessment attribute corresponding to the target seaborne cargo based on the product of the risk assessment attribute to be adjusted and the target correction coefficient; and determining the cargo risk level corresponding to the target seaborne cargo based on the risk assessment attribute range corresponding to the target risk assessment attribute.

[0049] The target risk assessment attribute can be understood as the final risk assessment attribute obtained after comprehensively assessing the risk of the target maritime cargo using risk assessment dimensions and risk influencing factors. The risk assessment attribute range can be understood as a pre-defined range used to match and determine the risk level of the target maritime cargo.

[0050] For example, four risk assessment attribute ranges are preset, and the target risk attribute is denoted as S. When S≥75, the target sea freight corresponds to the first cargo risk level; when 50≤S≤74, the target sea freight corresponds to the second cargo risk level; when 25≤S≤49, the target sea freight corresponds to the third cargo risk level; and when S<25, the target sea freight corresponds to the fourth cargo risk level.

[0051] In this technical solution, the higher the target risk assessment attribute, the higher the risk level and degree of risk of the target maritime cargo. To promptly alert relevant users to the target maritime cargo, warnings are issued based on the pre-set warning methods corresponding to each risk assessment attribute range. For example, when the target maritime cargo corresponds to the first risk level, a red warning message prompts the user to conduct a manual review of the target maritime cargo; when the target maritime cargo corresponds to the second risk level, a yellow warning message prompts the user to independently assess whether to conduct a review; when the target maritime cargo corresponds to the third risk level, a blue warning message prompts the user to conduct routine monitoring of the target maritime cargo; and when the target maritime cargo corresponds to the fourth risk level, the user can be prompted to release the target maritime cargo.

[0052] In practical applications, warning methods can include flashing lights, sound alerts, text messages, or emails. This technical solution does not limit the warning method.

[0053] In this technical solution, the maritime cargo risk identification method based on dynamic weight and factor correction further includes: prioritizing risk detection of the cargo association information of the target maritime cargo based on at least one high-risk detection condition to obtain the detection result; if the cargo association information triggers any high-risk detection condition, then the cargo risk level of the target maritime cargo is determined to be a high-risk level.

[0054] The detection result is either a hit or a miss. Cargo-related information may include the target seaborne cargo's technical specifications, customer authorization status, historical booking comparison information, and shipper-related information.

[0055] In practical applications, priority is given to identifying high-risk targets for maritime cargo. Once the preset high-risk detection conditions are triggered, the target maritime cargo is directly classified as high-risk, without needing to initiate a comprehensive assessment across any risk assessment dimension. This setup prioritizes the identification and handling of the most serious safety hazards. On one hand, it relies on independent and rigid high-risk rules to form a safety net, unaffected by model biases in the confidence levels and dynamic weighting calculations of various assessment dimensions. This avoids underestimating extreme risks through multi-dimensional comprehensive scoring, preventing the release of high-risk cargo and effectively mitigating major property safety accidents. On the other hand, the judgment result can be quickly output through simple rule matching, eliminating the need for complex calculations such as retrieving multi-dimensional data, statistically calculating confidence levels, dynamically adjusting weights, and comprehensively calculating risk scores. This significantly shortens the response time for cargo risk identification, adapts to real-time screening and inspection scenarios for large volumes of cargo, and allows for the pre-diversion of high-risk and ordinary risk cargo. Differentiated and rapid management of cargo at different risk levels can be achieved without waiting for complete comprehensive assessment results, comprehensively balancing risk control reliability, system operating efficiency, and business processing convenience.

[0056] The technical solution of this invention involves obtaining the risk assessment attributes of a target maritime cargo under at least one risk assessment dimension and determining the weights to be used for each risk assessment dimension; determining the risk assessment attributes to be adjusted for the target maritime cargo based on the risk assessment attributes associated with each risk assessment dimension and the weights to be used; obtaining a target correction coefficient based on the risk factors corresponding to at least one risk influencing factor associated with the target maritime cargo; and determining the cargo risk level of the target maritime cargo based on the risk assessment attributes to be adjusted and the target correction coefficient. This technical solution first determines the risk assessment attributes to be adjusted for the target maritime cargo under multiple risk assessment dimensions, then combines the risk detection conditions corresponding to the risk influencing factors that may be triggered during the actual transportation process of the target maritime cargo, and calculates the target correction coefficient based on the risk factors corresponding to the triggered risk influencing factors. Based on this, the target correction coefficient is used to correct the parameters of the risk assessment attributes to be adjusted to obtain the target risk assessment attributes corresponding to the target maritime cargo, and then the cargo risk level of the target maritime cargo is determined based on the risk assessment attribute range to which the target risk assessment attributes belong. This solves the problem in the prior art of accurately determining the cargo risk level of cargo, and achieves the effect of accurately determining the cargo risk level of the target maritime cargo by combining multiple risk assessment dimensions and risk influencing factors.

[0057] Example 2 Figure 2The flowchart of a method for identifying maritime cargo risks based on dynamic weights and factor correction provided in Embodiment 2 of the present invention includes, optionally, the following steps before obtaining the risk assessment attributes of the target maritime cargo under at least one risk assessment dimension and determining the weights to be used for each risk assessment dimension: obtaining the average risk prediction accuracy rate of all risk assessment dimensions within a historical risk prediction period, and the first risk prediction accuracy rate and the first risk prediction false alarm rate for each risk assessment dimension; wherein, the historical risk prediction period is at least one preceding period of the current risk prediction period; and updating the current weights of each risk assessment dimension in the historical risk prediction period based on the average risk prediction accuracy rate, each first risk prediction accuracy rate, and each first risk prediction false alarm rate, to obtain the weights to be used for each risk assessment dimension.

[0058] like Figure 2 As shown, the method includes: S210. Obtain the average risk prediction accuracy rate of all risk assessment dimensions within the historical risk prediction period, as well as the first risk prediction accuracy rate and the first risk prediction false alarm rate for each risk assessment dimension.

[0059] The historical risk prediction period is at least one preceding period of the current risk prediction period. It is used to periodically review the risk level assessment results of target maritime cargoes and dynamically adjust the weights of each risk assessment dimension over a specific time period. For example, each quarter, the current weights of each risk assessment dimension are adjusted to obtain usable weights, which are then used as the usable weights for the corresponding risk assessment dimensions in the next quarter. In this case, the risk prediction period is a quarterly period, and the historical risk prediction period is the preceding period adjacent to the current risk prediction period.

[0060] The first risk prediction accuracy rate characterizes the probability of accurate risk prediction for a single risk assessment dimension of the target maritime cargo within the historical risk prediction period. The first risk prediction false alarm rate characterizes the probability of false alarm for a single risk assessment dimension of the target maritime cargo within the historical risk prediction period. The average risk prediction accuracy rate is determined based on the average of all first risk prediction accuracy rates.

[0061] S220. Based on the average risk prediction accuracy, the accuracy of each first risk prediction, and the false alarm rate of each first risk prediction, update the current weight of each risk assessment dimension in the historical risk prediction period to obtain the weight to be used for each risk assessment dimension.

[0062] The current weight refers to the weight of each risk assessment dimension within the historical risk prediction period.

[0063] In this technical solution, each risk assessment dimension is configured with dynamically adjustable weights to be used. The core function of these weights is to indicate that the higher the weight of a risk assessment dimension, the greater the contribution of its output data to the overall risk assessment result of the target maritime cargo. However, in practical applications, the predictive accuracy of different risk assessment dimensions varies across different scenarios or cargo samples. It is possible that some risk assessment dimensions may misjudge or miss risks, yet still maintain a high weight, directly lowering the overall accuracy of risk classification.

[0064] Based on this, this paper introduces the average risk prediction accuracy rate within the historical risk prediction period, the prediction accuracy rate of each first risk, and the false alarm rate of each first risk prediction to dynamically adjust the current weight of each risk assessment dimension, so as to obtain the weight to be used for each risk assessment dimension.

[0065] The advantage of this setting is that by dynamically adjusting the weights of each risk assessment dimension, the influence of each risk assessment dimension on the final target risk assessment attribute can be adaptively allocated, avoiding the problem that fixed weights cannot adapt to the differences in the accuracy of dimension predictions, and ultimately significantly improving the accuracy of the risk level determination of the target maritime cargo.

[0066] Optionally, based on the average risk prediction accuracy, the accuracy of each first risk prediction, and the false alarm rate of each first risk prediction, the current weight of each risk assessment dimension in the historical risk prediction period is updated to obtain the weight to be used for each risk assessment dimension. This includes: selecting first risk assessment dimensions that meet the first condition from all risk assessment dimensions and increasing the current weight of the first risk assessment dimension to obtain the first weight to be optimized; selecting second risk assessment dimensions that meet the second condition from all risk assessment dimensions and decreasing the current weight of the second risk assessment dimension to obtain the second weight to be optimized; selecting third risk assessment dimensions that do not meet the first condition and do not meet the second condition from all risk assessment dimensions and using the current weight of the third risk assessment dimension as the third weight to be optimized; and normalizing each first weight to be optimized, each second weight to be optimized, and each third weight to be optimized to obtain the weight to be used for each risk assessment dimension.

[0067] The first condition is that the accuracy rate of the first risk prediction is higher than the accuracy rate of the average risk prediction, and the difference between the accuracy rate of the first risk prediction and the accuracy rate of the average risk prediction is greater than a first difference. The second condition is that the false alarm rate of the first risk prediction is greater than the preset false alarm rate for at least two consecutive historical risk prediction periods.

[0068] Optionally, if a new risk assessment dimension is added, the current weights of all risk assessment dimensions are normalized to obtain the weights to be used for each risk assessment dimension.

[0069] For example, taking a quarterly cycle as an example, the accuracy rate of each risk assessment dimension in all historical cargo risk assessment cases in the current period (i.e., the accuracy rate of the first risk prediction) is calculated. If the accuracy rate of the first risk prediction of a certain risk assessment dimension is higher than the average accuracy rate of the risk prediction, and the first difference between the accuracy rate of the first risk prediction and the average accuracy rate of the risk prediction is greater than the first difference (i.e., the first condition is met), then the current weight of that risk assessment dimension is increased by 0.05, with an upper limit of 0.35.

[0070] If the false alarm rate of the first risk prediction of a certain risk assessment dimension is greater than the preset false alarm rate (e.g., the preset false alarm rate is 20%) for at least two consecutive historical risk prediction periods, then the weight to be used for that risk assessment dimension will be reduced by 0.05, with a lower limit of 0.1.

[0071] To ensure the consistency of the overall weights, after updating each risk assessment dimension, the weights to be used for each risk assessment dimension are normalized to ensure that the sum of the total weights to be used is 1. Furthermore, when adding a new risk assessment dimension, the corresponding weights are allocated to each risk assessment dimension proportionally to the initial weights of each risk assessment dimension.

[0072] The advantage of this setup is that, since the weights of the risk assessment dimensions are in a relative proportion, unilaterally adjusting the weights of only some dimensions can disrupt the original proportional balance. This can lead to a situation where the actual influence of the risk assessment dimension with increased weight may actually decrease due to the relative passive dilution of the weights of other risk assessment dimensions, or the actual influence of the dimension with decreased weight may increase due to the relative passive increase of the weights of other dimensions. Therefore, this paper normalizes the adjusted weights of all dimensions after adjusting the weights of some dimensions to eliminate the distortion of relative influence caused by the imbalance in proportions. This ensures that the actual risk assessment influence of each dimension is strictly consistent with the set weights, thereby improving the accuracy of the cargo risk level assessment of the target maritime cargo.

[0073] S230. Obtain the risk assessment attributes of the target seaborne cargo under at least one risk assessment dimension, and determine the weights to be used for each risk assessment dimension.

[0074] S240. Based on the risk assessment attributes and weights to be used associated with each risk assessment dimension, determine the risk assessment attributes to be adjusted for the target seaborne cargo.

[0075] S250. Based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo, the target correction coefficient is obtained.

[0076] S260. Based on the risk assessment attributes to be adjusted and the target correction coefficient, determine the cargo risk level corresponding to the target seaborne cargo.

[0077] The technical solution of this invention, before obtaining the risk assessment attributes of the target maritime cargo under at least one risk assessment dimension and determining the weights to be used for each risk assessment dimension, further includes: obtaining the risk prediction confidence level of each risk assessment dimension for the target maritime cargo within the historical cargo transportation cycle; and updating the current weight of each risk assessment dimension based on the current risk prediction confidence level to obtain the weights to be used. This technical solution dynamically adjusts the weights to be used for each risk assessment dimension based on the risk prediction results of each risk assessment dimension for the target maritime cargo, so as to adaptively adjust the influence ratio of each risk assessment dimension on the assessment of the cargo risk level when actually calculating the cargo risk level of the target maritime cargo, thereby achieving a more accurate assessment of the cargo risk level of the target maritime cargo.

[0078] Example 3 Figure 3 This is a schematic diagram of a maritime cargo risk identification system based on dynamic weights and factor correction, provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes: a weight determination module 310, a risk assessment attribute determination module 320, a target correction coefficient determination module 330, and a cargo risk level determination module 340.

[0079] The weight determination module 310 is used to obtain the risk assessment attributes of the target sea freight under at least one risk assessment dimension, and determine the weights to be used for each risk assessment dimension. The risk assessment attribute determination module 320 is used to determine the risk assessment attributes to be adjusted for the target seaborne cargo based on the risk assessment attributes associated with each risk assessment dimension and the weights to be used. The target correction coefficient determination module 330 is used to obtain the target correction coefficient based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo. The cargo risk level determination module 340 is used to determine the cargo risk level corresponding to the target seaborne cargo based on the risk assessment attributes to be adjusted and the target correction coefficient.

[0080] The technical solution of this invention involves obtaining the risk assessment attributes of a target maritime cargo under at least one risk assessment dimension and determining the weights to be used for each risk assessment dimension; determining the risk assessment attributes to be adjusted for the target maritime cargo based on the risk assessment attributes associated with each risk assessment dimension and the weights to be used; obtaining a target correction coefficient based on the risk factors corresponding to at least one risk influencing factor associated with the target maritime cargo; and determining the cargo risk level of the target maritime cargo based on the risk assessment attributes to be adjusted and the target correction coefficient. This technical solution first determines the risk assessment attributes to be adjusted for the target maritime cargo under multiple risk assessment dimensions, then combines the risk detection conditions corresponding to the risk influencing factors that may be triggered during the actual transportation process of the target maritime cargo, and calculates the target correction coefficient based on the risk factors corresponding to the triggered risk influencing factors. Based on this, the target correction coefficient is used to correct the parameters of the risk assessment attributes to be adjusted to obtain the target risk assessment attributes corresponding to the target maritime cargo, and then the cargo risk level of the target maritime cargo is determined based on the risk assessment attribute range to which the target risk assessment attributes belong. This solves the problem in the prior art of accurately determining the cargo risk level of cargo, and achieves the effect of accurately determining the cargo risk level of the target maritime cargo by combining multiple risk assessment dimensions and risk influencing factors.

[0081] Optionally, the maritime cargo risk identification system based on dynamic weights and factor correction further includes: an acquisition module, used to acquire, before acquiring the risk assessment attributes of the target maritime cargo under at least one risk assessment dimension and determining the weights to be used for each risk assessment dimension, the average risk prediction accuracy rate of all said risk assessment dimensions within a historical risk prediction period, and the first risk prediction accuracy rate and the first risk prediction false alarm rate for each risk assessment dimension; wherein, the historical risk prediction period is at least one preceding period of the current risk prediction period; The weight determination module is used to update the current weight of each risk assessment dimension in the historical risk prediction period based on the average risk prediction accuracy, the accuracy of each of the first risk predictions, and the false alarm rate of each of the first risk predictions, so as to obtain the weight to be used for each of the risk assessment dimensions.

[0082] Optionally, the weight determination module includes: a first weight determination unit for optimization, used to select a first risk assessment dimension that meets a first condition from all the risk assessment dimensions, and increase the current weight of the first risk assessment dimension to obtain a first weight for optimization; wherein, the first condition is that the first risk prediction accuracy is higher than the average risk prediction accuracy, and the difference between the first risk prediction accuracy and the average risk prediction accuracy is greater than a first difference. The second unit for determining the weight to be optimized is used to select a second risk assessment dimension that meets the second condition from all the risk assessment dimensions, and reduce the current weight of the second risk assessment dimension to obtain the second weight to be optimized; wherein, the second condition is that the false alarm rate of the first risk prediction is greater than the preset false alarm rate for at least two consecutive historical risk prediction periods. The third weight determination unit is used to filter out the third risk assessment dimensions that do not meet the first condition and do not meet the second condition from all the risk assessment dimensions, and to take the current weight of the third risk assessment dimension as the third weight to be optimized. The weight determination unit is used to normalize each of the first weights to be optimized, each of the second weights to be optimized, and each of the third weights to be optimized to obtain the weights to be used corresponding to each of the risk assessment dimensions.

[0083] Optionally, the maritime cargo risk identification system based on dynamic weights and factor correction is also used to normalize the current weights of all the risk assessment dimensions if there are new risk assessment dimensions, so as to obtain the weights to be used for each of the risk assessment dimensions.

[0084] Optionally, the target correction coefficient determination module includes: a first factor determination unit, used to determine the first factor based on the product of the risk factors of all risk factors if the target seaborne cargo triggers risk detection conditions corresponding to at least two risk influencing factors; The upper limit risk factor determination unit is used to dynamically determine the upper limit risk factor based on the factor correlation between at least two risk influencing factors. The first target correction coefficient determination unit is used to determine the first factor as the target correction coefficient if the first factor is less than or equal to the upper limit risk factor. The second target correction coefficient determination unit is used to determine the upper limit risk factor as the target correction coefficient if the first factor is greater than the upper limit risk factor.

[0085] Optionally, the target correction coefficient determination module further includes: a value to be used determination unit, used to determine the correlation coefficient between each pair of risk influencing factors, and to determine the value to be used based on all correlation coefficients; The effective factor number determination unit is used to determine the number of effective factors based on the value to be used and the first preset function; The factor dynamic determination unit is used to dynamically determine the upper limit risk factor based on the number of effective factors and a second preset function.

[0086] Optionally, the cargo risk level determination module includes: a target risk assessment attribute determination unit, used to determine the target risk assessment attribute corresponding to the target seaborne cargo based on the product of the risk assessment attribute to be adjusted and the target correction coefficient; The cargo risk level determination unit is used to determine the cargo risk level of the target seaborne cargo based on the risk assessment attribute range corresponding to the target risk assessment attribute.

[0087] Optionally, the maritime cargo risk identification system based on dynamic weights and factor correction further includes: a detection result determination module, used to perform risk detection on the cargo association information of the target maritime cargo based on at least one high-risk detection condition, and obtain the detection result; wherein, the detection result is either a hit or a miss; The high-risk level determination module is used to determine the risk level of the target seaborne cargo as high-risk if the cargo-related information triggers any high-risk detection condition.

[0088] The maritime cargo risk identification system based on dynamic weight and factor correction provided in the embodiments of the present invention can execute the maritime cargo risk identification method based on dynamic weight and factor correction provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0089] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and no limitation is imposed herein.

[0090] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for identifying maritime cargo risks based on dynamic weights and factor corrections, characterized in that, include: Obtain the risk assessment attributes of the target seaborne cargo under at least one risk assessment dimension, and determine the weights to be used for each risk assessment dimension; Based on the risk assessment attributes and weights to be used associated with each of the aforementioned risk assessment dimensions, the risk assessment attributes to be adjusted for the target seaborne cargo are determined. Based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo, the target correction coefficient is obtained; Based on the risk assessment attributes to be adjusted and the target correction coefficient, the cargo risk level corresponding to the target seaborne cargo is determined.

2. The method according to claim 1, characterized in that, Before obtaining the risk assessment attributes of the target maritime cargo under at least one risk assessment dimension and determining the weights to be used for each risk assessment dimension, the method further includes: Obtain the average risk prediction accuracy rate of all risk assessment dimensions within the historical risk prediction period, as well as the first risk prediction accuracy rate and the first risk prediction false alarm rate for each risk assessment dimension; wherein, the historical risk prediction period is at least one preceding period of the current risk prediction period; Based on the average risk prediction accuracy, the accuracy of each of the first risk predictions, and the false alarm rate of each of the first risk predictions, the current weight of each of the risk assessment dimensions in the historical risk prediction period is updated to obtain the weight to be used for each of the risk assessment dimensions.

3. The method according to claim 2, characterized in that, The step of updating the current weights of each risk assessment dimension in the historical risk prediction period based on the average risk prediction accuracy, the accuracy of each of the first risk predictions, and the false alarm rate of each of the first risk predictions, to obtain the weights to be used for each of the risk assessment dimensions, includes: From all the risk assessment dimensions, a first risk assessment dimension that meets the first condition is selected, and the current weight of the first risk assessment dimension is increased to obtain the first weight to be optimized; wherein, the first condition is that the first risk prediction accuracy is higher than the average risk prediction accuracy, and the difference between the first risk prediction accuracy and the average risk prediction accuracy is greater than the first difference. From all the risk assessment dimensions, a second risk assessment dimension that meets the second condition is selected, and the current weight of the second risk assessment dimension is reduced to obtain the second weight to be optimized; wherein, the second condition is that the false alarm rate of the first risk prediction is greater than the preset false alarm rate for at least two consecutive historical risk prediction periods; From all the risk assessment dimensions, a third risk assessment dimension that does not meet the first condition and does not meet the second condition is selected, and the current weight of the third risk assessment dimension is taken as the third weight to be optimized. Normalize each of the first, second, and third weights to be optimized to obtain the weights to be used for each of the risk assessment dimensions.

4. The method according to claim 2, characterized in that, Also includes: If a new risk assessment dimension is added, the current weights of all the risk assessment dimensions are normalized to obtain the weights to be used for each risk assessment dimension.

5. The method according to claim 1, characterized in that, The process of determining the risk assessment attributes to be adjusted for the target maritime cargo based on the risk assessment attributes and weights associated with each of the aforementioned risk assessment dimensions includes: For each of the aforementioned risk assessment dimensions, the first attribute is obtained based on the risk assessment attributes associated with the current risk assessment dimension and the weights to be used; The first attribute corresponding to all risk assessment dimensions is superimposed to obtain the risk assessment attribute to be adjusted for the target seaborne cargo.

6. The method according to claim 1, characterized in that, The target correction coefficient is obtained based on the risk factor corresponding to at least one risk influencing factor associated with the target maritime cargo, including: If the target seaborne cargo triggers risk detection conditions corresponding to at least two of the risk influencing factors, then the first factor is determined based on the product of the risk factors of all the risk influencing factors. The upper limit risk factor is dynamically determined based on the factor correlation between at least two of the aforementioned risk influencing factors. If the first factor is less than or equal to the upper limit risk factor, then the first factor is used as the target correction coefficient; If the first factor is greater than the upper limit risk factor, then the upper limit risk factor is used as the target correction coefficient.

7. The method according to claim 6, characterized in that, The step of dynamically determining the upper limit risk factor based on the factor correlation between at least two of the risk influencing factors includes: Determine the correlation coefficients between each pair of risk influencing factors, and determine the values ​​to be used based on all correlation coefficients; Based on the value to be used and the first preset function, determine the number of effective factors; The upper limit risk factor is dynamically determined based on the number of effective factors and the second preset function.

8. The method according to claim 1, characterized in that, Determining the cargo risk level corresponding to the target maritime cargo includes: The target risk assessment attribute corresponding to the target seaborne cargo is determined based on the product of the risk assessment attribute to be adjusted and the target correction coefficient. Based on the risk assessment attribute range corresponding to the target risk assessment attribute, the cargo risk level corresponding to the target maritime cargo is determined.

9. The method according to claim 1, characterized in that, Also includes: Risk detection is performed on the cargo association information of the target maritime cargo based on at least one high-risk detection condition to obtain a detection result; wherein the detection result is either a hit or a miss. If the cargo association information triggers any of the high-risk detection conditions, then the cargo risk level of the target seaborne cargo is determined to be high-risk.

10. A maritime cargo risk identification system based on dynamic weights and factor correction, characterized in that, include: The weight determination module is used to obtain the risk assessment attributes of the target seaborne cargo under at least one risk assessment dimension, and determine the weights to be used for each risk assessment dimension. The risk assessment attribute determination module is used to determine the risk assessment attribute to be adjusted for the target seaborne cargo based on the risk assessment attributes associated with each of the risk assessment dimensions and the weights to be used. The target correction coefficient determination module is used to obtain the target correction coefficient based on the risk factor corresponding to at least one risk influencing factor associated with the target seaborne cargo; The cargo risk level determination module is used to determine the cargo risk level corresponding to the target seaborne cargo based on the risk assessment attribute to be adjusted and the target correction coefficient.

Images