System of long-term prediction for each company

Abstract

A method for shipping pattern generation, which comprises clustering, by a processor, shipping records associated with a plurality of parties and generating standard shipping patterns associated with the plurality of parties based on the clustered shipping records; loading, by the processor, shipping records and standard shipping pattern associated with a target party from the plurality of parties; performing, by the processor, data augmentation to correct data points of the target party's shipping records using the standard shipping patterns associated with the plurality of parties and updating the target party's standard shipping pattern based on the corrected data points; training, by the processor, a prediction model for performing shipping trend prediction of the target party, wherein training of the prediction model is performed using the target party's standard shipping pattern; deploying the trained prediction model.

Description

SYSTEM OF LONG-TERM PREDICTION FOR EACH COMPANYBACKGROUNDField

[0001] The present disclosure is generally directed to a method and a system for shipping pattern generation.Related Art

[0002] A product is composed of various components, which are also composed of various parts and subcomponents. The process through which the product is supplied is the supply chain. In the supply chain, shipping, especially maritime shipping, has significant and direct impact on product’s availability and pricing. Therefore, monitoring of shipping risks and preparing countermeasures against such risks are essential for a business’ continuity. Many enterprise shipping monitoring services are available, with some providing the added feature of future shipping status prediction. Such future shipping status prediction services may include Estimated Time of Arrival (ETA) prediction, congestion prediction at cargo ports, and cargo accumulation prediction at cargo ports. Machine learning (ML) may be utilized in realizing these predictions. More specifically, ML extracts time-series trend features from past shipping records as training data and models shipping trends with extracted time-series trend features.

[0003] In the related art, a future shipping delay prediction method utilizing learned shipping trends per shipping carrier is disclosed. While the method provides delay prediction per shipping carrier, it fails to provide shipping trend analysis / prediction and model generation per shipping company. A shipping company is a business / company that utilizes the shipping service of a shipping carrier. Model training through ML involves shipping features extraction by eliminating the outliers and extracting shipping trends. Feature extraction is realized by finding the shipping patterns from a sufficient amount of data to discover the patterns. Thus, model generation on the per shipping company becomes difficult when there is insufficient amount of data on the shipping companies’ shipping records.

[0004] Therefore, a new method / solution that enables performance of long-term shipping trend prediction per shipping company is needed.SUMMARY

[0005] Aspects of the present disclosure involve an innovative method for shipping pattern generation. The method may include clustering, by a processor, shipping records associated with a plurality of parties and generating standard shipping patterns associated with the plurality of parties based on the clustered shipping records; loading, by the processor, shipping records and standard shipping pattern associated with a target party from the plurality of parties; performing, by the processor, data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties and updating the target party’s standard shipping pattern based on the corrected data points; training, by the processor, a prediction model for performing shipping trend prediction of the target party, wherein training of the prediction model is performed using the target party’s standard shipping pattern; deploying the trained prediction model; and entering a future date as input into the trained prediction model and receiving predicted shipping trend of the target party associated with the future date as output from the trained prediction model.

[0006] Aspects of the present disclosure involve an innovative non-transitory computer readable medium, storing instructions for shipping pattern generation. The instructions may include clustering, by a processor, shipping records associated with a plurality of parties and generating standard shipping patterns associated with the plurality of parties based on the clustered shipping records; loading, by the processor, shipping records and standard shipping pattern associated with a target party from the plurality of parties; performing, by the processor, data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties and updating the target party’s standard shipping pattern based on the corrected data points; training, by the processor, a prediction model for performing shipping trend prediction of the target party, wherein training of the prediction model is performed using the target party’s standard shipping pattern; deploying the trained prediction model; and entering a future date as input into the trained prediction model and receiving predicted shipping trend of the target party associated with the future date as output from the trained prediction model.

[0007] Aspects of the present disclosure involve an innovative server system for performing shipping pattern generation. The system may include a memory; and a processor in communicationwith the memory, the processor is configured to: cluster shipping records associated with a plurality of parties and generate standard shipping patterns associated with the plurality of parties based on the clustered shipping records, and store the standard shipping patterns in the memory, wherein the shipping records are stored in the memory; load shipping records and standard shipping pattern associated with a target party from the plurality of parties; perform data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties and update the target party’s standard shipping pattern based on the corrected data points; train a prediction model for performing shipping trend prediction of the target party, wherein training of the prediction model is performed using the target party’s standard shipping pattern; deploy the trained prediction model; and enter a future date as input into the trained prediction model and receive predicted shipping trend of the target party associated with the future date as output from the trained prediction model.

[0008] Aspects of the present disclosure involve an innovative system for shipping pattern generation. The system may include means for clustering shipping records associated with a plurality of parties and generating standard shipping patterns associated with the plurality of parties based on the clustered shipping records, means for loading shipping records and standard shipping pattern associated with a target party from the plurality of parties; means for performing data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties and updating the target party’s standard shipping pattern based on the corrected data points; means for training a prediction model for performing shipping trend prediction of the target party, wherein training of the prediction model is performed using the target party’s standard shipping pattern; means for deploying the trained prediction model; and means for entering a future date as input into the trained prediction model and receiving predicted shipping trend of the target party associated with the future date as output from the trained prediction model.BRIEF DESCRIPTION OF DRAWINGS

[0009] A general architecture that implements the various features of the disclosure will now be described with reference to the drawings. The drawings and the associated descriptions are provided to illustrate example implementations of the disclosure and not to limit the scope of thedisclosure. Throughout the drawings, reference numbers are reused to indicate correspondence between referenced elements.

[0010] FIG. 1 illustrates an example long-term prediction system 100, in accordance with an example implementation.

[0011] FIG. 2 illustrates an example process flow 200 of the long-term prediction system 100, in accordance with an example implementation.

[0012] FIG. 3 illustrates an example process 300 for performing extraction of standard shipping patterns, in accordance with an example implementation.

[0013] FIG. 4 illustrates an example process 400 for performing data augmentation, in accordance with an example implementation.

[0014] FIG. 5 illustrates examples of shipping pattern clustering and data point augmentation, in accordance with an example implementation.

[0015] FIG. 6 illustrates an alternative example long-term prediction system 600, in accordance with an example implementation.

[0016] FIG. 7 illustrates an example user interface 700 for contract checking using long-term shipping trend prediction, in accordance with an example implementation.

[0017] FIG. 8 illustrates an alternative example long-term prediction system 800, in accordance with an example implementation.

[0018] FIG. 9 illustrates an example process flow 900 for performing event impact estimation, in accordance with an example implementation.

[0019] FIG. 10 illustrates an example user interface 1000 for contract checking using event- considered long-term shipping trend prediction, in accordance with an example implementation.

[0020] FIG. 11 illustrates an example computing environment with an example computing device suitable for use in some example implementations.DETAILED DESCRIPTION

[0021] The following detailed description provides details of the figures and example implementations of the present application. Reference numerals and descriptions of redundant elements between figures are omitted for clarity. Terms used throughout the description are provided as examples and are not intended to be limiting. For example, the use of the term “automatic” may involve fully automatic or semi-automatic implementations involving user or administrator control over certain aspects of the implementation, depending on the desired implementation of one of the ordinary skills in the art practicing implementations of the present application. Selection can be conducted by a user through a user interface or other input means, or can be implemented through a desired algorithm. Example implementations as described herein can be utilized either singularly or in combination, and the functionality of the example implementations can be.

[0022] Example implementations utilize Machine Learning (ML) / Artificial Intelligence (Al) in predicting goods shipping trends of companies and their value accumulation associated with each company in a given time and geolocation. More particularly, example implementations relate to utilization of ML / AI systems in predicting long-term shipping value accumulation associated with each company at a designated / identified port.

[0023] Example implementations involve a method and system that perform long-term shipping trend prediction per shipping company. Example implementations perform: (1) standard shipping patterns extraction from shipping records of all shipping companies; and (2) data augmentation for each company following standard shipping pattern extraction.

[0024] Standard shipping patterns represent an annual increase or decrease in shipping volume or value. Such changes may include, increase in shipping volume during the months in summer, decrease in shipping volume during the months in winter, shipping volume fluctuation based on occurrence of an event, etc. These pattems / trends are closely associated with products that each company handles. Example implementations take company shipping records as time-series data and apply time-series clustering to produce representative shipping patterns as outputs. Shipping companies having similar shipping trends are grouped together with their shipping recordsclustered together as a representative shipping pattern. These shipping patterns represent standard shipping patterns for the different shipping record groups.

[0025] Clustering indicates the group and the standard shipping pattern to which each company belongs. Since the standard shipping pattern is known, outliers can be identified and removed through standard shipping pattern comparison, and data augmentation can be utilized to outer / interpolate missing data generated from the removal.

[0026] Shipping companies that have only began their trades tend to have few shipping records. Example implementations provide shipping trend estimation for newly formed / junior shipping companies through generation of a shipping trend prediction model and predict shipping trends derived from a similarly situated shipping trend group. Example implementations search for shipping companies that share commonalities with the newly formed / junior shipping company which may include products handled, transaction value, location, clients, ports, and other information, and predict the future shipping trends to which the newly formed / junior shipping company belongs.

[0027] FIG. 1 illustrates an example long-term prediction system 100, in accordance with an example implementation. The long-term prediction system 100 may include components such as, but not limited to, a shipping records crawler 110, a shipping records database 120, a standard shipping clustering module 130, a standard shipping patterns database 140, a data augmentation module 150, an Artificial Intelligence (Al) model training module 160, a company shipping prediction model database 170, a model searcher module 180, and a model loader & prediction module 190.

[0028] The shipping records crawler 110 performs the function of shipping records extraction / crawling. The long-term prediction system 100 communicates with data storages of the various shipping companies and storages of third-party data / service providers in possession of shipping companies’ data through network 10, where data and information associated with each shipping company are received by the shipping records crawler 110 through the network 10 for performance of shipping records extraction / crawling. In some example implementations, shipping records crawling is performed at predetermined or arbitrary intervals, such as a week, a month,etc. In some other example implementations, the period for performing shipping records crawling can adjusted by a user of the system.

[0029] The extracted / crawled shipping records are then stored in the shipping records database 120 as time-series data. In some example implementations, the shipping records are indexed according to name of shipping company, consignee company, shipping item, item value, port of destination, port of origin, departure date, arrival date, etc.

[0030] The standard shipping clustering module 130 performs clustering of shipping records to generate standard shipping patterns. The shipping records of the shipping companies are retrieved from the shipping records database 120 by the standard shipping clustering module 130 to perform time-series clustering. Clustering is performed by classifying similar shipping trends into the same class / group and generating standard shipping pattern for each class / group. Each standard shipping pattern represents the shipping trends classified into the same class / group. The generated standard shipping patterns are then stored in the standard shipping patterns database 140.

[0031] Data augmentation module 150 then retrieves the standard shipping patterns stored in the standard shipping patterns database 140 for processing. Specifically, the outlier removal module 151 of the data augmentation module 150 is used to identify and remove outliers from shipping records associated with a company through use of associated standard shipping pattern. In addition, interpolation module 152 is used to outer / interpolate any void, whether originally existing or generated through the outlier removal module 151.

[0032] The Al model training module 160 utilizes the augmented shipping records / data points to train a long-term shipping prediction model associated with the selected / target company. The trained long-term shipping prediction model is then stored in the company shipping prediction model database 170.

[0033] Training of the long-term prediction model may be performed using at least one of a Machine Learning (ML) / Artificial Intelligence method or a statistical analysis method. The ML / AI method may utilize one of long short-term memory (LSTM), convolutional neural networks (CNN), or any other machine learning network for training the long-term prediction model.Specifically, a model is learnt that predicts shipping trend based on input information such as future data, shipping company data, and arrival port information. The generated shipping trend may include the amount value of the shipping items.

[0034] The statistical analysis method may utilize any regression analysis method in derivation of coefficients of the explanatory variables from which the object variable is derived. Specifically, explanatory variables may include, but not limited to, the seasonality, the moving average, the autoregression, and the trend.

[0035] The model searcher module 180 receives input / user request from the input terminal 20 and searches for a model stored in the company shipping prediction model database 170 that corresponds to the input. The input terminal 20 may be any user input device such as, but not limited to, a keyboard, a mouse, touch screen, or other user interface to provide input that is associated with a computing device such as a desktop computer, laptop computer, mobile device, etc.

[0036] Once a model is detected from the company shipping prediction model database 170, the model is then loaded to the model loader & prediction module 190 where long-term shipping prediction is performed based on the received input from the input terminal 20. The generated long-term shipping prediction is then sent to the output terminal 30 for display and user review. A user requests the long-term prediction by inputting the date, the shipping company, and the arrival port. After that, the user can view the shipping trend and the amount value of items on the output terminal 30.

[0037] FIG. 2 illustrates an example process flow 200 of the long-term prediction system 100, in accordance with an example implementation. The process begins at step S202 where shipping records of shipping companies are extracted / crawled. At step S204, the crawled shipping records are stored in the shipping records database 120.

[0038] At step S206, the system generates standard shipping patterns by clustering all shipping records stored in the shipping records database 120. Clustering is performed by classifying similar shipping trends into the same class. A standard shipping pattern represents the shipping trends classified into the same class, which will be described in more detail below. At step S208, thesystem loads the shipping records of a model target company selected from the shipping companies and identifies the target company’s standard shipping pattern.

[0039] The process then continues to step S210, where outliers are detected / identified and removed from the shipping records of the target company using the standard shipping pattern associated with the target company. Outering / interpolating the missing value is then performed using the standard shipping pattern associated with the target company at step S212. Specifically, the missing value is interpolated to resemble the standard shipping pattern and the result is denoted as augmented shipping records.

[0040] At step S214, a prediction model is trained using the augmented shipping records, and stored in the company shipping prediction model database 170 on completion of the training at step S216. The prediction model is a Machine Learning (ML) / Artificial Intelligence (AI) model that predicts long-term shipping trend / pattem associated with a selected company.

[0041] The long-term prediction system 100 thus identifies the shipping pattern for each company by arrival port and predicts the long-term shipping trend and the amount value of items for each company by the arrival port. Through use of long-term prediction, a user can estimate the amount of trade, future monetary risk, and value accumulation risk associated with a particular company, and make business decisions based on the estimates.

[0042] FIG. 3 illustrates an example process 300 for performing extraction of standard shipping patterns, in accordance with an example implementation. The process 300 is performed by the standard shipping clustering module 130 and the standard shipping patterns database 140. The process begins at step S302 where all shipping records are loaded by the standard shipping clustering module 130 from the shipping records database 120.

[0043] The process then proceeds to step S304 where shipping amount is aggregated on a predetermined basis (e.g. daily, weekly, etc.) by shipping company, arrival ports, and year intervals. The shipping amount may be in monetary unit or the number of cargoes. In some example implementations, the shipping amount may also be aggregated from other perspectives, such as consignee company, transshipment port, etc.

[0044] At step S306, data points are defined as annual shipping trends and clustering is performed for all data points such that data points of shipping companies with similar trends are classified into the same class. FIG. 5 illustrates examples of shipping pattern clustering and data point augmentation, in accordance with an example implementation. As illustrated in FIG. 5, data points associated with the three companies of “AAA Company”, “BBB Company”, and “CCC Company” are defined and divided on year intervals that span from 2020 to 2022. For example, data points 502 represent AAA Company’s data points for the period of 2020 associated with the Port XXX. Normalization may be performed to align the maximum values of all data points to measure trend similarities without being affected by differences in the total amount.

[0045] At step S308, center points are extracted from the different classes, wherein center point of a class may include mean or median of all data point contained in the same class. The center point of each class serves as representative data point for each class. In some example implementations, this center point is the data point with minimum sum of dynamic time warping distance to all data belonging to the same class. At step S310, the standard shipping clustering module 130 stores the standard shipping patterns, each aggregated data point, and data points belonging to class information in the standard shipping patterns database 140.

[0046] FIG. 4 illustrates an example process 400 for performing data augmentation, in accordance with an example implementation. The process 400 is performed by the data augmentation module 150, the outlier removal module 151, and the interpolation module 152. The process begins at step S402 where all shipping records and standard shipping patterns are loaded from the shipping records database 120 and the standard shipping patterns database 140 respectively.

[0047] The process then continues to step S404 where an arrival port is selected by the data augmentation module 150. The arrival port is selected from all ports contained in the shipping records. The order of the port selection can be arbitrary. At step S406, the data augmentation module 150 selects one shipping company that sends its items to the selected arrival port. The order of the shipping company selection can be arbitrary.

[0048] At step S408, the data augmentation module 150 gets the standard shipping pattern that matches closest with the yearly shipping records of the selected port and shipping company. Atstep S410, the outlier removal module 151 removes data point(s) that appear as outliers through comparison of yearly shipping records against the standard shipping pattern identified as most similar. As shown in FIG. 5, for the arrival port of Port XXX, data points of the shipping company BBB Company during the year of 2021 are identified as containing outliers. The outliers are removed from the data points through use of generated standard shipping pattern.

[0049] At step S412 the interpolation module 152 interpolates the yearly shipping records where data point is missing / removed so that trend conforms with the most similar standard shipping pattern. As shown in FIG. 5, for the arrival port of Port XXX, data points of the shipping company BBB Company during the year of 2021 that have been removed due to outlier removal are interpolated such that the shipping trend matches the standard shipping pattern.

[0050] At step S414, a determination is made as to whether all shipping companies with the selected arrival port have been processed. If the answer is no, then the process returns to step S406 where additional processing is performed for shipping companies that have yet been processed. If the answer is yes at step S414, then the process continues to step S416 where a determination is made as to whether all arrival ports have been processed / selected. If the answer is no, then the process returns to step S404 where additional processing is performed for arrival ports that have yet been processed / selected. If the answer is yes at step S4146 then the process comes to an end.

[0051] FIG. 6 illustrates an alternative example long-term prediction system 600, in accordance with an example implementation. The long-term prediction system 600 may include components of the long-term prediction system 100, such as the shipping records crawler 110, the shipping records database 120, the standard shipping clustering module 130, the standard shipping patterns database 140, the data augmentation module 150, the Artificial Intelligence (AI) model training module 160, the company shipping prediction model database 170, the model searcher module 180, and the model loader & prediction module 190. However, the long-term prediction system 600 may further include components such as, but not limited to, a contract parameter database 610 and a contract draft generator & checker 620.

[0052] When making insurance contracts with each company, it is possible to make appropriate premium quotations based on the shipping trend predicted for each company. The long-term prediction system 600 checks contracts through the contract parameter database 610 andthe contract draft generator & checker 620 using the generated long-term shipping trend predictions. With the long-term prediction system 600, a user, such as an insurer, can identify the potential risks and their amount values for each company. The user can then set the appropriate premium in contracting with each company.

[0053] For example, if the user, an insurer, makes a cargo insurance policy with AAA Company, the long-term prediction system 600 can be used to predict AAA Company’s shipping trend during the contract period. Specifically, the contract draft generator & checker 620 finds insurer risks, for example, high-value accumulation risk at certain ports or areas, etc. Once a risk is identified, the contract draft generator & checker 620 then loads the contract parameters, such as rates for premium calculation from the contract parameter database 610. The contract draft generator & checker 620 then calculates the appropriate premiums based on the contract parameters.

[0054] FIG. 7 illustrates an example user interface 700 for contract checking using long-term shipping trend prediction, in accordance with an example implementation. The user interface 700 is a graphical user interface (GUI) that comprises an input area 710, a contract suggestion area 720, a risk list 730, and a shipping trend prediction area 740. The input area 710 allows a company name and the contract period to be entered by a user. The contract suggestion area 720 provides information on suggested contract terms. The risk list 730 derived by the contract draft generator & checker 620 and identifies potential risks based on the generated long-term shipping trend predictions. The shipping trend prediction area 740 shows the shipping trend prediction for the input shipping company.

[0055] FIG. 8 illustrates an alternative example long-term prediction system 800, in accordance with an example implementation. The long-term prediction system 800 is similar to the long-term prediction system 600 of FIG. 6, but includes additional components such as an event database 810 and an event impact estimation module 820.

[0056] Predictions are subject to change with the occurrence of various events. For example, the shipping trend for each company is subject due to occurrence of extreme weather, canal accidents, port congestions, and changes in demand and supply. By taking the impacts of such events into consideration, a user of the long-term prediction system 800 is able to determine acontract's validity. Specifically, the long-term prediction system 800 calculates impacts from past events associated with acquired shipping records and performs prediction change simulations. With the long-term prediction system 800, a user, more particularly an insurer, can identify the impact of potential events on shipping trends and set the appropriate contract premium according to the estimated impact.

[0057] The event database 810 stores information / data on events, which are used for processing by the event impact estimation module 820. Event impact estimation module 820 estimates past event impact by comparing the shipping records against the standard shipping pattern. Using the estimated impact, event impact estimation module 820 shows the estimated impact if the event occurs during the prediction period.

[0058] FIG. 9 illustrates an example process flow 900 for performing event impact estimation, in accordance with an example implementation. The process begins at step S902 where the event impact estimation module 820 loads the standard shipping patterns and all data points generated from all shipping records from the standard shipping patterns database 140. At step S904, the event impact estimation module 820 loads an event and its accrual period from the event database 810.

[0059] Next, at step S906, the event impact estimation module 820 selects an arrival port. The arrival port is selected from all arrival ports contained in the shipping records. In some example implementations, the order of selection is arbitrary. At step S908, the event impact estimation module 820 selects a shipping company that sends its items to the selected arrival port. In some example implementations, the order of selection is arbitrary.

[0060] The process then continues to step S910 where the event impact estimation module 820 checks if the data of the selected port and shipping company for that period differ from the standard shipping pattern. At step S912, a determination is made as to whether difference is detected in the comparison.

[0061] If the answer is no, then the process proceeds to step S916. If the answer is yes, then the process continues to step S914 where the event impact estimation module 820 calculates the difference in percentages and stores the result along with the event, the selected port, and the selected shipping company to the event database 810.

[0062] At step S916, a determination is made as to whether all companies with the selected arrival port have been processed. If the answer is no, then the process returns to step S908 and reselects another company for processing. If the answer is yes, then the process continues to step S918. At step S918, a determination is made as to whether all ports included in the shipping records have been selected and processed. If the answer is no, then the process returns to step S906 and re-selects another port included in the shipping records for processing. If the answer is yes, then the process comes to an end.

[0063] FIG. 10 illustrates an example user interface 1000 for contract checking using event- considered long-term shipping trend prediction, in accordance with an example implementation. The user interface 1000 is similar to the user interface 700 of FIG. 7, but with the addition of an events & month area 1010. The events & month area 1010 allows the user to select registered events stored in the event database 810 and the estimated month in performing long-term shipping trend prediction to estimate even impact on shipping. The user interface 1000 further includes an estimated loss column 1020 in the risk list 730, which provides estimated maximum loss calculated based impact of the selected event. In addition, the impact of the selected event 1030 is shown and illustrated in the shipping trend prediction area 740.

[0064] The foregoing example implementation may have various benefits and advantages. For example, identifying shipping pattern for a target company by arrival port, and predicting the longterm shipping trend and values of items associated with each company through information of the arrival port. In addition, prediction is achieved even when a target company has insufficient shipping records for performing machine learning or statistical analysis. Through use of long-term prediction, a user can estimate the amount of trade, future monetary risk, and value accumulation risk for a particular company, and make business decisions based on the estimates.

[0065] FIG. 11 illustrates an example computing environment with an example computing device suitable for use in some example implementations. Computing device 1105 in computing environment 1100 can include one or more processing units, cores, or processors) 1110, memory 1115 (e.g., RAM, ROM, and / or the like), internal storage 1120 (e.g., magnetic, optical, solid-state storage, and / or organic), and / or I / O interface 1125, any of which can be coupled on a communication mechanism or bus 1130 for communicating information or embedded in thecomputing device 1105. I / O interface 1125 is also configured to receive images from cameras or provide images to projectors or displays, depending on the desired implementation.

[0066] Computing device 1105 can be communicatively coupled to inpuVuser interface 1135 and output device / interface 1140. Either one or both of the input / user interface 1135 and output device / interface 1140 can be a wired or wireless interface and can be detachable. Input / user interface 1135 may include any device, component, sensor, or interface, physical or virtual, that can be used to provide input (e.g., buttons, touch-screen interface, keyboard, a pointing / cursor control, microphone, camera, braille, motion sensor, accelerometer, optical reader, and / or the like). Output device / interface 1140 may include a display, television, monitor, printer, speaker, braille, or the like. In some example implementations, input / user interface 1135 and output device / interface 1140 can be embedded with or physically coupled to the computing device 1105. In other example implementations, other computing devices may function as or provide the functions of input / user interface 1135 and output device / interface 1140 for a computing device 1105.

[0067] Examples of computing device 1105 may include, but are not limited to, highly mobile devices (e.g., smartphones, devices in vehicles and other machines, devices carried by humans and animals, and the like), mobile devices (e.g., tablets, notebooks, laptops, personal computers, portable televisions, radios, and the like), and devices not designed for mobility (e.g., desktop computers, other computers, information kiosks, televisions with one or more processors embedded therein and / or coupled thereto, radios, and the like).

[0068] Computing device 1105 can be communicatively coupled (e.g., via I / O interface 1125) to external storage 1145 and network 1150 for communicating with any number of networked components, devices, and systems, including one or more computing devices of the same or different configuration. Computing device 1105 or any connected computing device can be functioning as, providing services of, or referred to as, a server, client, thin server, general machine, special-purpose machine, or another label.

[0069] I / O interface 1125 can include, but is not limited to, wired and / or wireless interfaces using any communication or I / O protocols or standards (e.g., Ethernet, 802.11x, Universal System Bus, WiMax, modem, a cellular network protocol, and the like) for communicating information toand / or from at least all the connected components, devices, and network in computing environment 1100. Network 1150 can be any network or combination of networks (e.g., the Internet, local area network, wide area network, a telephonic network, a cellular network, satellite network, and the like).

[0070] Computing device 1105 can use and / or communicate using computer-usable or computer-readable media, including transitory media and non-transitory media. Transitory media include transmission media (e.g., metal cables, fiber optics), signals, carrier waves, and the like. Non-transitory media include magnetic media (e.g., disks and tapes), optical media (e.g., CD ROM, digital video disks, Blu-ray disks), solid-state media (e.g., RAM, ROM, flash memory, solid-state storage), and other non-volatile storage or memory.

[0071] Computing device 1105 can be used to implement techniques, methods, applications, processes, or computer-executable instructions in some example computing environments. Computer-executable instructions can be retrieved from transitory media, and stored on and retrieved from non-transitory media. The executable instructions can originate from one or more of any programming, scripting, and machine languages (e.g., C, C++, C#, Java, Visual Basic, Python, Perl, JavaScript, and others).

[0072] Processors) 1110 can execute under any operating system (OS) (not shown), in a native or virtual environment. One or more applications can be deployed that include logic unit 1160, application programming interface (API) unit 1165, input unit 1170, output unit 1175, and inter-unit communication mechanism 1195 for the different units to communicate with each other, with the OS, and with other applications (not shown). The described units and elements can be varied in design, function, configuration, or implementation and are not limited to the descriptions provided. Processors) 1110 can be in the form of hardware processors such as central processing units (CPUs) or in a combination of hardware and software units.

[0073] In some example implementations, when information or an execution instruction is received by API unit 1165, it may be communicated to one or more other units (e.g., logic unit 1160, input unit 1170, output unit 1175). In some instances, logic unit 1160 may be configured to control the information flow among the units and direct the services provided by API unit 1165, the input unit 1170, and the output unit 1175 in some example implementations described above.For example, the flow of one or more processes or implementations may be controlled by logic unit 1160 alone or in conjunction with API unit 1165. The input unit 1170 may be configured to obtain input for the calculations described in the example implementations, and the output unit 1175 may be configured to provide an output based on the calculations described in example implementations.

[0074] Processors) 1110 can be configmed to cluster shipping records associated with a plurality of parties and generating standard shipping patterns associated with the plurality of parties based on the clustered shipping records as illustrated in FIG. 2. The processor(s) 1110 may also be configmed to load shipping records and standard shipping pattern associated with a target party from the plurality of parties as illustrated in FIG. 2. The processor(s) 1110 may also be configmed to perform data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties and updating the target party’s standard shipping pattern based on the corrected data points as illustrated in FIG. 2. The processors) 1110 may also be configmed to train a prediction model for performing shipping trend prediction of the target party, wherein training of the prediction model is performed using the target party’s standard shipping pattern as illustrated in FIG. 2. The processors) 1110 may also be configmed to deploy the trained prediction model as illustrated in FIG. 2. The processors) 1110 may also be configmed to enter a future date as input into the trained prediction model and receiving predicted shipping trend of the target party associated with the future date as output from the trained prediction model as illustrated in FIG. 2.

[0075] The processors) 1110 may also be configured to load the shipping records associated with the plurality of parties as illustrated in FIGS. 8-10. The processor s) 1110 may also be configured to load an event and accrual period associated with the event from an event database as illustrated in FIGS. 8-10. The processors) 1110 may also be configured to identify an arrival port, wherein the arrival port is associated with shipping records of a first party of the plurality of parties as illustrated in FIGS. 8-10. Theprocessor(s) 1110 may also be configured to load port data associated with the arrival port as illustrated in FIGS. 8-10. The processor(s) 1110 may also be configured to compare the port data and the shipping records of the first party against standard shipping pattern associated with the first party during the accrual period as illustrated in FIGS. 8- 10. The processors) 1110 may also be configmed for a difference being detected from thecomparison, calculate percentage difference between the port data and the standard shipping pattern associated with the first party during the accrual period and percentage difference between the shipping records of the first party and the standard shipping pattern associated with the first party during the accrual period as illustrated in FIGS. 8-10. The processors) 1110 may also be configured to display the percentage differences on a graphical user interface (GUI) as illustrated in FIGS. 8-10.

[0076] The processors) 1110 may also be configured to identify insurance risk based on the predicted shipping trend of the target party as illustrated in FIGS. 8-10. The processor(s) 1110 may also be configured to, for insurance risk being identified, calculate an insurance premium to be added to an insurance contract associated with the target party based on the identified insurance risk and updating the insurance contract to include the insurance premium as illustrated in FIGS. 8-10. The processor(s) 1110 may also be configured to display the predicted shipping trend of the target party on a graphical user interface (GUI) as illustrated in FIGS. 5, 7, and 10.

[0077] Some portions of the detailed description are presented in terms of algorithms and symbolic representations of operations within a computer. These algorithmic descriptions and symbolic representations are the means used by those skilled in the data processing arts to convey the essence of their innovations to others skilled in the art. An algorithm is a series of defined steps leading to a desired end state or result. In example implementations, the steps carried out require physical manipulations of tangible quantities for achieving a tangible result.

[0078] Unless specifically stated otherwise, as apparent from the discussion, it is appreciated that throughout the description, discussions utilizing terms such as “processing,” “computing,” “calculating,” “determining,” “displaying,” or the like, can include the actions and processes of a computer system or other information processing device that manipulates and transforms data represented as physical (electronic) quantities within the computer system’s registers and memories into other data similarly represented as physical quantities within the computer system’s memories or registers or other information storage, transmission or display devices.

[0079] Example implementations may also relate to an apparatus for performing the operations herein. This apparatus may be specially constructed for the required purposes, or it may include one or more general-purpose computers selectively activated or reconfigured by one or morecomputer programs. Such computer programs may be stored in a computer readable medium, such as a computer readable storage medium or a computer readable signal medium. A computer readable storage medium may involve tangible mediums such as, but not limited to, optical disks, magnetic disks, read-only memories, random access memories, solid-state devices and drives, or any other types of tangible or non-transitory media suitable for storing electronic information. A computer readable signal medium may include mediums such as carrier waves. The algorithms and displays presented herein are not inherently related to any particular computer or other apparatus. Computer programs can involve pure software implementations that involve instructions that perform the operations of the desired implementation.

[0080] Various general-purpose systems may be used with programs and modules in accordance with the examples herein, or it may prove convenient to construct a more specialized apparatus to perform desired method steps. In addition, the example implementations are not described with reference to any particular programming language. It will be appreciated that a variety of programming languages may be used to implement the teachings of the example implementations as described herein. The instructions of the programming language(s) may be executed by one or more processing devices, e.g., central processing units (CPUs), processors, or controllers.

[0081] As is known in the art, the operations described above can be performed by hardware, software, or some combination of software and hardware. Various aspects of the example implementations may be implemented using circuits and logic devices (hardware), while other aspects may be implemented using instructions stored on a machine-readable medium (software), which if executed by a processor, would cause the processor to perform a method to carry out implementations of the present application. Further, some example implementations of the present application may be performed solely in hardware, whereas other example implementations may be performed solely in software. Moreover, the various functions described can be performed in a single unit, or can be spread across a number of components in any number of ways. When performed by software, the methods may be executed by a processor, such as a general-purpose computer, based on instructions stored on a computer readable medium. If desired, the instructions can be stored in the medium in a compressed and / or encrypted format.

[0082] Moreover, other implementations of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the teachings of the present application. Various aspects and / or components of the described example implementations may be used singly or in any combination. It is intended that the specification and example implementations be considered as examples only, with the true scope and spirit of the present application being indicated by the following claims.

Claims

CLAIMSWhat is claimed is:

1. A method for shipping pattern generation, the method comprising: clustering, by a processor, shipping records associated with a plurality of parties and generating standard shipping patterns associated with the plurality of parties based on the clustered shipping records; loading, by the processor, shipping records and standard shipping pattern associated with a target party from the plurality of parties; performing, by the processor, data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties and updating the target party’s standard shipping pattern based on the corrected data points; training, by the processor, a prediction model for performing shipping trend prediction of the target party, wherein training of the prediction model is performed using the target party’s standard shipping pattern; deploying the trained prediction model; and entering a future date as input into the trained prediction model and receiving predicted shipping trend of the target party associated with the future date as output from the trained prediction model.

2. The method of claim 1 , wherein clustering shipping records associated with the plurality of parties is performed on arbitrary intervals.

3. The method of claim 1 , wherein the prediction model is trained using an artificial intelligence (Al) model or statistical model.

4. The method of claim 1 , wherein generating the standard shipping patterns based on the clustered shipping records comprises: loading the shipping records associated with the plurality of parties; aggregating, for each party of the plurality of parties, shipping volumes by at least one of arrival ports or predetermined time unit, wherein shipping volumes are derived from shipping records associated with respective party;defining, for each party of the plurality of parties, the aggregated shipping volumes as a plurality of annual shipping trends; and clustering, for each party of the plurality of parties, the plurality of annual shipping trends into associated standard shipping pattern.

5. The method of claim 1 , wherein performing data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties comprises: identifying at least one party other than the target party from the plurality of parties having a first shipping information in common with the target party, wherein the first shipping information is contained in shipping records of the at least one party and the target party; identifying a standard shipping pattern associated with the at least one party that matches closest to the target party’s shipping records; and performing data augmentation to correct data points of the target party’s shipping records based on the identified standard shipping pattern.

6. The method of claim 5, wherein identifying the standard shipping pattern associated with the at least one party that matches closest to the target party’s shipping records is based yearly shipping records of the target party’s shipping records.

7. The method of claim 1, wherein data augmentation comprises at least one of outlier removal or missing value interpolation on data points of the target party’s shipping records.

8. The method of claim 1, further comprising: loading, by the processor, the shipping records associated with the plurality of parties; loading, by the processor, an event and accrual period associated with the event from an event database; identifying an arrival port, wherein the arrival port is associated with shipping records of a first party of the plurality of parties; loading, by the processor, port data associated with the arrival port;comparing, by the processor, the port data and the shipping records of the first party against standard shipping pattern associated with the first party during the accrual period; for a difference being detected from the comparison, calculating percentage difference between the port data and the standard shipping pattern associated with the first party during the accrual period and percentage difference between the shipping records of the first party and the standard shipping pattern associated with the first party during the accrual period; and displaying the percentage differences on a graphical user interface (GUI).

9. The method of claim 1 , further comprising: identifying, by the processor, insurance risk based on the predicted shipping trend of the target party; and for insurance risk being identified, calculating, by the processor, an insurance premium to be added to an insurance contract associated with the target party based on the identified insurance risk and updating the insurance contract to include the insurance premium.

10. The method of claim 1, further comprising: displaying the predicted shipping trend of the target party on a graphical user interface (GUI).

11. A system for performing shipping pattern generation, the system comprising: a memory; and a processor in communication with the memory, the processor is configured to: cluster shipping records associated with a plurality of parties and generate standard shipping patterns associated with the plurality of parties based on the clustered shipping records, and store the standard shipping patterns in the memory, wherein the shipping records are stored in the memory; load shipping records and standard shipping pattern associated with a target party from the plurality of parties; perform data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties and update the target party’s standard shipping pattern based on the corrected data points;train a prediction model for performing shipping trend prediction of the target party, wherein training of the prediction model is performed using the target party’s standard shipping pattern; deploy the trained prediction model; and enter a future date as input into the trained prediction model and receive predicted shipping trend of the target party associated with the future date as output from the trained prediction model.

12. The system of claim 11, wherein cluster shipping records associated with the plurality of parties is performed on arbitrary intervals.

13. The system of claim 11, wherein the prediction model is trained using an artificial intelligence (Al) model or statistical model.

14. The system of claim 11, wherein the processor is configured to generate the standard shipping patterns based on the clustered shipping records by: loading the shipping records associated with the plurality of parties from the memory; aggregating, for each party of the plurality of parties, shipping volumes by at least one of arrival ports or predetermined time unit, wherein shipping volumes are derived from shipping records associated with respective party; defining, for each party of the plurality of parties, the aggregated shipping volumes as a plurality of annual shipping trends; and clustering, for each party of the plurality of parties, the plurality of annual shipping trends into associated standard shipping pattern.

15. The system of claim 11, wherein the processor is configured to perform data augmentation to correct data points of the target party’s shipping records using the standard shipping patterns associated with the plurality of parties by: identifying at least one party other than the target party from the plurality of parties having a first shipping information in common with the target party, wherein the first shipping information is contained in shipping records of the at least one party and the target party;identifying a standard shipping pattern associated with the at least one party that matches closest to the target party’s standard shipping pattern; and performing data augmentation to correct data points of the target party’s shipping records based on the identified standard shipping pattern.

16. The system of claim 15, wherein identify the standard shipping pattern associated with the at least one party that matches closest to the target party’s shipping records is based yearly shipping records of the target party’s shipping records.

17. The system of claim 11, wherein data augmentation comprises at least one of outlier removal or missing value interpolation on data points of the target party’s shipping records.

18. The system of claim 11, wherein the processor is further configured to: load the shipping records associated with the plurality of parties from the memory; load an event and accrual period associated with the event from an event database; identify an arrival port, wherein the arrival port is associated with shipping records of a first party of the plurality of parties; loading port data associated with the arrival port; comparing the port data and the shipping records of the first party against standard shipping pattern associated with the first party during the accrual period; for a difference being detected from the comparison, calculate percentage difference between the port data and the standard shipping pattern associated with the first party during the accrual period and percentage difference between the shipping records of the first party and the standard shipping pattern associated with the first party during the accrual period; and display the percentage differences on a graphical user interface (GUI).

19. The system of claim 11, wherein the processor is further configured to: identify insurance risk based on the predicted shipping trend of the target party; andfor insurance risk being identified, calculate an insurance premium to be added to an insurance contract associated with the target party based on the identified insurance risk and update the insurance contract to include the insurance premium.

20. The system of claim 11, further comprising: a graphical user interface (GUI), and wherein the processor is further configured to display the predicted shipping trend of the target party on the GUI.

Images