Method, device, equipment, medium and product for predicting service handling waiting time

Abstract

This invention discloses a method, apparatus, device, medium, and product for predicting business processing waiting time, relating to the fields of big data and artificial intelligence technology, specifically the application of large-scale models in the fintech field. The method includes: identifying other users whose business processing order precedes that of the target user, identifying at least one candidate business step included in the pending business of these other users, and determining a baseline time interval corresponding to each candidate business step; determining a time correction coefficient based on the profiles of other users and the current teller's profile, and correcting the baseline time interval based on the time correction coefficient to determine the predicted time for each business step corresponding to the candidate business step; determining the total predicted time for the pending business based on the predicted time for each business step, and determining the business processing waiting time for the target user based on the total predicted time. This invention improves the accuracy of business processing waiting time prediction.

Description

Technical Field

[0001] This invention relates to the fields of big data and artificial intelligence, specifically to the application of large models in the field of financial technology, and particularly to a method, apparatus, device, medium, and product for predicting business processing waiting time. Background Technology

[0002] In the daily operations of bank branches, the difficulty in accurately estimating customer wait times for transactions has become a core pain point affecting customer experience and banking service efficiency. Transaction wait time refers to the duration required for a customer to begin their transaction at a bank branch.

[0003] Traditional waiting time prediction methods typically rely on a simple average of historical transaction processing times; however, this method suffers from inaccurate waiting time predictions. Summary of the Invention

[0004] This invention provides a method, apparatus, equipment, medium, and product for predicting business processing waiting time, in order to solve the problem of inaccurate prediction in traditional business processing waiting time prediction.

[0005] According to one aspect of the present invention, a method for predicting business processing waiting time is provided, the method comprising:

[0006] Identify at least one other user whose business processing order is ahead of the target user, and identify at least one candidate business step included in the pending business of the other user, and determine the baseline time interval corresponding to the candidate business step;

[0007] Based on the user profiles of other users and the current teller profile, a time-consuming correction coefficient is determined, and the baseline time-consuming interval is corrected according to the time-consuming correction coefficient to determine the predicted time of the business process corresponding to the candidate business process; wherein, the current teller is the teller who handles the pending business.

[0008] The total predicted time for the pending business is determined based on the predicted time for each business process, and the waiting time for the target user is determined based on the total predicted time for each business process.

[0009] According to another aspect of the present invention, a business processing waiting time prediction device is provided, the device comprising:

[0010] The information determination module is used to determine at least one other user whose business processing order is ahead of the target user, and to determine at least one candidate business step included in the pending business of the other user, and to determine the baseline time interval corresponding to the candidate business step;

[0011] The business process time prediction module is used to determine the time correction coefficient based on the other user profiles of the other users and the current teller profile of the current teller, and to correct the baseline time interval based on the time correction coefficient to determine the predicted time of the business process corresponding to the candidate business process; wherein, the current teller is the teller who handles the pending business.

[0012] The business processing waiting time prediction module is used to determine the total predicted time of the pending business based on the predicted time of the business process, and to determine the business processing waiting time for the target user based on the total predicted time of the business process.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] At least one processor; and

[0015] A memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the business processing waiting time prediction method according to any one of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the business processing waiting time prediction method according to any one of the present invention.

[0018] According to another aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the business processing waiting time prediction method according to any one of the present invention.

[0019] This invention identifies at least one other user whose business processing order precedes that of the target user, determines at least one candidate business step included in the pending business of the other user, and determines a baseline time interval corresponding to the candidate business step; based on the profiles of the other users and the current teller profile, it determines a time correction coefficient, and corrects the baseline time interval according to the time correction coefficient to determine the predicted time of the business step corresponding to the candidate business step; wherein, the current teller is the teller handling the pending business; based on the predicted time of the business step, it determines the total predicted time of the pending business, and based on the total predicted time of the business, it determines the business processing waiting time for the target user. The beneficial effects are:

[0020] This invention fully considers the differences in complexity of different candidate business processes (reflected in the baseline time interval), the teller profile characteristics of different tellers, and the user profile characteristics of different users in the process of predicting business processing wait times. Compared with the traditional method that relies on the simple averaging of historical business processing times, this invention can improve the accuracy of business processing wait time prediction.

[0021] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments 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 these drawings without creative effort.

[0023] Figure 1 A flowchart illustrating a method for predicting business processing waiting time according to Embodiment 1 of the present invention;

[0024] Figure 2 A flowchart illustrating a method for predicting business processing waiting time according to Embodiment 2 of the present invention;

[0025] Figure 3 This is a schematic diagram of a business processing waiting time prediction device provided in Embodiment 3 of the present invention;

[0026] Figure 4 This is a schematic diagram of the structure of an electronic device that implements the business processing waiting time prediction method of this invention. Detailed Implementation

[0027] 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.

[0028] It should be noted that the terms "target," "other," "candidate," "first," "second," "pending," and "beginning of processing," etc., used 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 embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] Example 1

[0030] Figure 1 This is a flowchart illustrating a method for predicting business processing waiting times according to Embodiment 1 of the present invention. This embodiment is applicable to predicting the corresponding waiting times for users waiting to process business at bank branches. The method can be executed by a business processing waiting time prediction device, which can be implemented in hardware and / or software, such as using a computer. Figure 1 As shown, the method includes:

[0031] S101. Identify at least one other user whose business processing order is ahead of the target user, and identify at least one candidate business step included in the pending business of the other user, and determine the baseline time interval corresponding to the candidate business step.

[0032] The "service processing order" refers to the priority ranking of users' transactions according to preset rules at a bank branch. For example, suppose there are five users—A, B, C, D, and E—who need to process transactions at a bank branch. The priority of processing these transactions will be ranked according to preset rules, such as the order: User A, User C, User E, User B, User D, etc. This example only explains the service processing order and does not specify a concrete order.

[0033] The target user refers to any user in a bank branch who has a predicted waiting time for their transaction. This can specifically refer to a particular user in the branch who needs to conduct business, or it can refer to any user in the branch who needs to conduct business. Other users refer to all users whose transaction processing order precedes that of the target user. For example, if the transaction processing order is user A, user C, user E, user B, and user D, where "user B" is the target user, then the other users include "user A," "user C," and "user E," a total of three users.

[0034] Pending transactions for other users refer to transactions submitted by other users that are in an intermediate state before the teller has completed their processing. It's understandable that each pending transaction is essentially process-driven, consisting of a series of interconnected stages or steps, known as "candidate transaction steps." Without at least one candidate transaction step, a pending transaction cannot operate or achieve its goal, reflecting that pending transactions are dynamic processes rather than static entities.

[0035] For example, suppose at a bank branch, another user's pending transaction is "opening a bank card account." This pending transaction could include, but is not limited to, "initial consultation," "document review and form completion," "identity verification and information entry," "password setting and card activation," and "certificate delivery and transaction confirmation," etc. This example only illustrates the potential transaction steps and does not specify any particular steps.

[0036] The baseline time interval for each candidate business step refers to the reasonable time range for each candidate business step, obtained by summarizing historical time data. This range is used to measure the processing efficiency of each candidate business step. It can be understood that the baseline time interval consists of a lower baseline time limit and a higher baseline time limit.

[0037] In one implementation, the current business processing order is obtained, and at least one other user whose business processing order is ahead of the target user is identified from the users who need to process the business.

[0038] After identifying other users, in one implementation, the business requirement description text of each other user is generated based on the collected voice descriptions of their business requirements. The target large language model is then used to analyze the business requirement description text and output at least one candidate business step included in the pending business of each other user.

[0039] After identifying other users, in another implementation, standardized process templates for different business types, such as bank account opening, transfer, and card loss reporting, are predefined. The corresponding candidate business steps are automatically matched based on the business type selected by each other user. For example, if the business type of any other user's pending transaction is "bank card loss reporting," the standardized process template is retrieved, and the candidate business steps included in the pending transaction are determined to be identity verification, filling out the loss reporting form, confirming the loss reporting information, and obtaining a temporary certificate.

[0040] After identifying other users, in another implementation, candidate business steps for each other user's pending business are dynamically generated using a decision tree model based on the business requirement characteristics submitted by each other user.

[0041] Furthermore, after determining the candidate business steps for each pending business, in one implementation, based on the historical time consumption of each candidate business step, the average historical time consumption and the standard deviation of historical time consumption for each candidate business step are determined, and based on the average historical time consumption and the standard deviation of historical time consumption, the baseline time consumption interval for each candidate business step is determined.

[0042] After identifying the candidate business steps for each pending business, in another implementation, domain experts decompose the candidate business steps according to the business process, set the theoretical time consumption weight of each candidate business step based on experience, and collect the historical time consumption of each candidate business step through simulation testing or sandbox environment. The initial weights are then calibrated using a regression model to generate the benchmark time consumption range for each candidate business step.

[0043] S102. Based on the user profiles of other users and the current teller profile of the current teller, determine the time consumption correction coefficient, and correct the baseline time consumption interval according to the time consumption correction coefficient to determine the predicted time consumption of the business process corresponding to the candidate business process.

[0044] Among them, "other user profiles" refers to the user profiles corresponding to other users. These user profiles are tagged user models abstracted from collected and analyzed multi-dimensional user data, used to accurately describe user characteristics and predict behavioral patterns. "Current teller profile" refers to the teller profile corresponding to the current teller. This profile is a tagged capability model built based on the teller's work performance, ability characteristics, and behavioral data, used to accurately assess the teller's comprehensive capabilities, optimize job matching, and improve service quality. The current teller is the teller handling pending transactions.

[0045] It is understandable that for the same candidate business process, different user profiles for other users, and different current teller profiles for the current teller, can lead to different processing times for the same candidate business process. For example, for candidate business process A, when other users correspond to other user profile 1 and the current teller corresponds to current teller profile 1, the processing time for candidate business process A is T1; when other users correspond to other user profile 2 and the current teller corresponds to current teller profile 2, the processing time for candidate business process A is T2, so there will be a discrepancy between T1 and T2.

[0046] Therefore, this embodiment of the invention determines a time-consuming correction coefficient based on other user profiles and the current teller profile. The time-consuming correction coefficient is a calibration parameter used to adjust the deviation between the baseline time interval and the predicted time for each business process, primarily to improve the accuracy of business process time prediction. The predicted time for each business process refers to the estimated processing time required for each candidate business process.

[0047] In one implementation, with the authorization of each other user and the current teller, the profiles of other users and the current teller are obtained.

[0048] Furthermore, in one implementation, based on the profiles of other users and the current teller profile, target transactions with the same profile combination are matched from the historical transactions. The ratio is calculated based on the first total historical time for each historical transaction and the second total historical time for each target transaction to determine the time correction coefficient.

[0049] In another implementation, a dynamic correction model is constructed based on historical behavior data and real-time status in other user profiles, combined with skill assessment indicators in the current teller profile. If the dynamic correction model determines that other user profiles show high-frequency trading habits and the current teller profile reflects excellent processing efficiency, a time reduction coefficient is assigned to the baseline time consumption coefficient; otherwise, a time expansion coefficient is assigned to obtain the time correction coefficient.

[0050] After determining the time consumption correction coefficient, the lower limit and upper limit of the baseline time consumption for each candidate business segment are determined based on the baseline time consumption interval of each candidate business segment. Based on the lower limit, upper limit and time consumption correction coefficient, each baseline time consumption interval is corrected to obtain the predicted time consumption of each candidate business segment.

[0051] It should be noted that the relevant information (including but not limited to other user profiles, current teller profiles, user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data comply with the relevant laws, regulations and standards of the relevant regions.

[0052] S103. Determine the total estimated processing time for the pending business based on the estimated processing time of each business step, and determine the processing waiting time for the target user based on the total estimated processing time.

[0053] The total predicted processing time refers to the total time required to complete each pending service. This is calculated by summing the predicted processing times for each candidate service step within each pending service. The processing waiting time refers to the waiting time required for a target user to begin processing the service.

[0054] In one implementation, the predicted time for each candidate business step within each pending business is summed to determine the total predicted time for each pending business. Further, the total predicted time for each pending business is summed to determine the sum of the predicted total times for all pending businesses. Then, based on the ratio of the sum of the predicted total times to the total number of tellers, the waiting time for the target user is determined.

[0055] This invention identifies at least one other user whose business processing order precedes that of the target user, determines at least one candidate business step included in the pending business of the other user, and determines a baseline time interval corresponding to the candidate business step; based on the profiles of the other users and the current teller profile, a time correction coefficient is determined, and the baseline time interval is corrected according to the time correction coefficient to determine the predicted time of the business step corresponding to the candidate business step; wherein, the current teller is the teller handling the pending business; the total predicted time of the pending business is determined based on the predicted time of the business steps, and the waiting time for the target user is determined based on the total predicted time of the business steps. The beneficial effects are:

[0056] This invention achieves a comprehensive assessment of the impact of varying complexity of different candidate business processes (reflected in a baseline time interval), teller profile characteristics of different tellers, and user profile characteristics of different users on business processing wait time prediction. Compared to the traditional method that relies on a simple average of historical business processing times, this invention improves the accuracy of business processing wait time prediction, thereby enhancing the user service experience.

[0057] Example 2

[0058] Figure 2 This is a flowchart of a method for predicting business processing waiting time according to Embodiment 2 of the present invention. This embodiment further optimizes and expands the above embodiment and can be combined with the various optional implementation methods described above. For example... Figure 2 As shown, the method includes:

[0059] S201. Identify at least one other user whose business processing order precedes that of the target user. Based on the collected voice descriptions of other users' business needs, generate a business need description text. Based on the business need description text, determine the business need keywords. Also, perform semantic role annotation on the business need description text to generate semantic role-annotated text.

[0060] Among them, the business requirement description speech refers to the specific business requirements, operational goals, or problem descriptions expressed orally by other users using natural language. The business requirement description speech is the direct carrier of user intent. The core objective of semantic role labeling is to identify the semantic relationships between predicates and surrounding components in the business requirement description text, and to label these components as specific semantic roles, thereby revealing the deep semantic structure of the business requirement description text.

[0061] In one implementation, a preset speech recognition method is used to perform speech recognition on the collected business requirement descriptions from other users, generating business requirement description text. For example, the collected business requirement descriptions from other users are input into an end-to-end speech recognition model. The end-to-end speech recognition model uses a self-attention mechanism to capture and process the features in the business requirement descriptions, thereby achieving accurate speech-to-text conversion and outputting business requirement description text.

[0062] Furthermore, a pre-defined keyword extraction method is used to extract keywords from the business requirement description text to determine the business requirement keywords. For example, a keyword extraction algorithm combining term frequency-inverse document frequency and a rule engine is used to extract key information such as business type and business attributes from the business requirement description text as business requirement keywords.

[0063] Furthermore, the business requirement description text is input into a pre-trained natural language understanding model to achieve semantic role labeling of the business requirement description text, and output the business requirement description text with semantic role labeling as semantic role labeled text.

[0064] S202. Based on the keywords of business requirements and the semantic role annotation text, determine at least one candidate business step included in the pending business of other users.

[0065] In one implementation, based on business requirement keywords and semantic role-labeled text, a target large language model is used to analyze and predict business processes, and output the candidate business processes included in the pending business of each other user.

[0066] By generating text describing business needs based on voice recordings of other users' business needs, identifying keywords based on the text, and performing semantic role labeling on the text to generate semantic role-labeled text, the system can determine at least one candidate business step for other users' pending business based on the keywords and semantic role-labeled text. The beneficial effects are:

[0067] Firstly, existing technologies rely on manual recording of business requirements, which is inefficient and prone to misinterpretation, leading to significant deviations in the final waiting time prediction. This embodiment utilizes intelligent voice interaction technology to automatically recognize the text describing business requirements, improving efficiency and the accuracy of the final waiting time prediction.

[0068] Secondly, by integrating speech recognition, keyword extraction, and semantic role labeling technologies, a three-level processing chain of "speech → text → structured semantics" is formed, breaking through the limitations of single text analysis.

[0069] Thirdly, semantic role labeling makes the business logic implicit in the business requirement description text explicit, which helps in the subsequent analysis and understanding of the target large language model.

[0070] Optionally, based on business requirement keywords and semantic role-labeled text, determine at least one candidate business step included in the pending business of other users, including:

[0071] Based on the keywords and semantic roles of the business requirements, generate a description text of the business to be processed; input the description text of the business to be processed into the target large language model, and output at least one candidate business step included in the business to be processed based on the description text of the business to be processed.

[0072] Among them, a large language model refers to a deep learning model with a massive number of parameters, capable of generating natural language text or understanding the meaning of language text. Large language models can handle various natural language tasks, such as text classification, question answering, and dialogue. Typically, the parameter scale of a large language model can even reach hundreds of billions. In this embodiment, the large language model, through learning from a large number of historical business process documents and operational specifications, possesses the ability to generate standardized flowcharts based on business requirements.

[0073] In one implementation, text is combined based on business requirement keywords and semantic role-labeled text to generate a description text of the business to be processed. This description text, along with prompt words, is input into a target large-scale language model. The prompt words guide the target large-scale language model to predict the business process based on the description text. The target large-scale language model generates a directed acyclic graph (DAG) containing process nodes and their dependencies, further identifying each process node in the DAG as a candidate business step. By generating the description text of the business to be processed based on business requirement keywords and semantic role-labeled text, and inputting this description text into the target large-scale language model, which then outputs at least one candidate business step included in the business to be processed, the beneficial effects are:

[0074] Firstly, business requirement keywords are anchored to the core business domain, and semantic role annotations analyze elements such as action subjects, objects, and conditions to form a machine-readable semantic graph.

[0075] Secondly, structured input improves the reasoning accuracy of the target large language model and avoids misjudgment of intent caused by colloquial expressions in traditional methods.

[0076] Thirdly, the target large language model is trained based on a massive business knowledge base, which can automatically complete the compliance candidate business links that are not explicitly stated by the user, thereby improving the comprehensiveness and accuracy of the prediction of candidate business links.

[0077] Optionally, the target large language model is generated as follows:

[0078] Based on the sample business requirement description text, determine the sample business requirement keywords and sample semantic role annotation text; based on the sample business requirement keywords and sample semantic role annotation text, generate sample business description text; obtain the business process tags corresponding to the sample business description text, and fine-tune the pre-trained large language model based on the sample business description text and business process tags to obtain the target large language model.

[0079] Among them, the sample business requirement description text refers to the original business requirement expression text proposed by users without structure processing, which is used as a training sample to fine-tune the pre-trained large language model.

[0080] Sample business requirement keywords are a set of terms or phrases extracted from the sample business requirement description text that can summarize the core business objectives or key elements. Sample semantic role annotation text refers to the semantic role annotation text generated by annotating the sample business requirement description text with semantic roles. The specific implementation methods for determining the sample business requirement keywords and sample semantic role annotation text can be found in the embodiments of this invention, and will not be repeated here.

[0081] Business process labels are used to mark the specific business process stage to which the sample business description text belongs. Essentially, they are used to guide the pre-trained large language model to understand the stage characteristics of the business scenario.

[0082] A pre-trained large language model refers to a general-purpose large language model that has been initially trained using massive amounts of general-purpose text data. Understandably, the pre-trained large language model is insufficient for predicting specific scenarios in business processes and requires fine-tuning.

[0083] In this embodiment, fine-tuning refers to the technical process of retraining a pre-trained large language model using sample business description text and business process labels to adapt it to the specific scenario of business process prediction. Its core logic is to transform general language capabilities into specialized business capabilities by adjusting model parameters.

[0084] In one implementation, sample business requirement keywords and sample semantic role annotation text are combined to generate sample business description text. A model fine-tuning method is then used to fine-tune the pre-trained large language model using the sample business description text and business process labels, and the fine-tuned model is used as the target large language model.

[0085] Specifically, the sample business description text is input into a pre-trained large language model. The pre-trained large language model outputs the predicted business steps included in the sample business based on the sample business description text. Then, a loss value is calculated based on the predicted business steps and business step labels. The model parameters of the pre-trained large language model are updated through backpropagation based on the loss value until the model parameters of the pre-trained large language model converge, resulting in the target large language model.

[0086] By determining sample business requirement keywords and sample semantic role annotation text based on the sample business requirement description text, and generating sample business description text based on the sample business requirement keywords and sample semantic role annotation text, the business process tags corresponding to the sample business description text are obtained. Based on the sample business description text and business process tags, the pre-trained large language model is fine-tuned to obtain the target large language model. The beneficial effect is that by fine-tuning the pre-trained large language model based on the sample business description text and business process tags, the effect of targeted fine-tuning of a general pre-trained large language model is achieved, making the fine-tuned target large language model adaptable to the specific scenario of business process prediction, thus ensuring the accuracy of business process prediction.

[0087] S203. Based on the business type of the business to be processed, other user profiles and the current teller profile, generate a business tag to be queried, match the business tag to be queried with the candidate business tags of historical processed businesses, and determine at least one target business to be processed from the historical processed businesses based on the matching results.

[0088] Among them, "historical transactions" refers to transactions that have already been completed. The candidate transaction tags for historical transactions are generated based on the transaction type, the teller profile of the teller who processed the historical transaction, and the user profile of the user to whom the historical transaction belongs.

[0089] In one implementation, the business type of the pending transaction is obtained, and a query business tag is generated by combining the business type, other user profiles, and the current teller profile. Further, the query business tag is matched with candidate business tags for each historical transaction, and at least one historical transaction whose candidate business tag matches the query business tag is selected as the target transaction.

[0090] For example, assuming the business tag to be queried is "Business Type 1; User Profile 1; Teller Profile 1", then at least one historical business with the same candidate business tag "Business Type 1; User Profile 1; Teller Profile 1" will be selected as the target business.

[0091] S204. Based on the historical time consumption of the candidate business process in each target business process, determine the historical average time consumption and the historical standard deviation of time consumption, and determine the benchmark time consumption interval based on the historical average time consumption and the historical standard deviation of time consumption.

[0092] Among them, historical processing time refers to the statistical data on the actual execution time of each candidate business step in the past for each target business process. Historical average processing time refers to the average of the historical processing times of each candidate business step. Historical standard deviation of processing time refers to the standard deviation of the historical processing times of each candidate business step.

[0093] In one implementation, the historical average time of each business step in each candidate business step is calculated to determine the historical average time, and the standard deviation of the historical average time of each business step in each candidate business step is calculated to determine the historical standard deviation. Further, a baseline upper limit for time is determined based on the sum of the historical average time and the historical standard deviation, and a baseline lower limit for time is determined based on the difference between the historical average time and the historical standard deviation. Finally, a baseline time interval is determined based on the baseline lower limit and the baseline upper limit for time.

[0094] By generating tags for the services to be queried based on the service type, other user profiles, and the current teller profile, and matching these tags with candidate service tags from historical transactions, at least one target service is identified from the historical transactions based on the matching results. Furthermore, based on the historical time consumption of each candidate service step within each target service step, the historical mean and standard deviation of the consumption time are determined, and a baseline consumption time interval is established using these historical mean and standard deviation. The beneficial effects are:

[0095] Firstly, based on two-way tag matching of user profiles and teller profiles, priority is given to recommending target business transactions with similar historical scenarios, thereby improving the rationality of the baseline time interval prediction.

[0096] Secondly, the historical time standard deviation reflects business volatility, and the benchmark time interval is set to cover most normal scenarios, avoiding expected deviations caused by fixed thresholds and improving the accuracy of benchmark time interval prediction.

[0097] Thirdly, the dynamically generated baseline time interval provides users with transparent waiting expectations, alleviating queuing anxiety.

[0098] S205. Determine the global average total time based on the first historical total time corresponding to each historical transaction, and determine the target average total time based on the second historical total time corresponding to each target transaction; determine the time correction coefficient based on the ratio between the target average total time and the global average total time.

[0099] The first historical total time refers to the total processing time for each historical transaction. The second historical total time refers to the total processing time for each target transaction. The global average total time is determined based on the average of the first historical total times. The target average total time is determined based on the average of the second historical total times. The time correction factor is used to measure the efficiency difference between the current processing scenario and the global average level.

[0100] By determining the global average total time based on the first historical total time for each historical transaction, and determining the target average total time based on the second historical total time for each target transaction, and then determining the time correction coefficient based on the ratio between the target average total time and the global average total time, the beneficial effects are:

[0101] Firstly, it enables the time consumption correction coefficient to measure the efficiency difference between the current processing scenario and the global average level, thereby improving the rationality of the time consumption correction coefficient.

[0102] Secondly, the global average total time consumption reflects the comprehensive service capability, while the target average total time consumption reflects the bottleneck of a specific business. Combining the two improves the accuracy of the time consumption correction coefficient.

[0103] S206. Determine the lower limit and upper limit of the benchmark time based on the benchmark time interval, and determine the benchmark time difference based on the difference between the upper limit and the lower limit of the benchmark time.

[0104] S207. Determine the baseline time sum based on the sum between the baseline time lower limit and the baseline time difference, and determine the predicted time of the business process based on the product of the baseline time sum, the time correction coefficient, and the random number coefficient.

[0105] In one implementation, the predicted time consumption of a business process is determined as follows:

[0106] t_N=min_t+(max_t-min_t)*random(0.5,1)*k

[0107] Where t_N represents the predicted time of the business process, min_t represents the lower limit of the baseline time, max_t represents the upper limit of the baseline time, (max_t-min_t) represents the difference between the baseline time and the baseline time, min_t+(max_t-min_t) represents the sum of the baseline time, k represents the time correction coefficient, and random(0.5,1) represents the random number coefficient between 0.5 and 1.

[0108] By determining the lower and upper limits of the baseline time based on the baseline time interval, and determining the baseline time difference based on the difference between the upper and lower limits; determining the baseline time sum based on the sum of the lower and baseline time differences; and determining the predicted time for the business process based on the product of the baseline time sum, the time correction coefficient, and the random number coefficient, the beneficial effects are:

[0109] Firstly, introducing random number coefficients can simulate the time fluctuations that exist in actual business processing, making the predicted time consumption of business processes more consistent with reality.

[0110] Secondly, the time consumption correction coefficient can be understood as a dynamic correction factor for the baseline time consumption interval, while the random number coefficient simulates uncontrollable factors. The product mechanism of the two achieves a weighted fusion of the baseline value and the actual scenario, avoiding the rigidity problem of static prediction.

[0111] S208. Identify the services that are in the "start processing" state from the pending services, and determine the start processing duration for each service.

[0112] "Started processing" refers to applications that have been received and entered the formal processing stage, but have not yet been completed. "Started processing duration" refers to the time interval from the start of processing to the current moment.

[0113] S209. Take the total predicted time for each business transaction as the target predicted total time, and determine the time deviation coefficient based on the start time and the target predicted total time. Determine the business processing waiting time for the target user based on the predicted total time, the total number of tellers, and the time deviation coefficient.

[0114] The total predicted time sum is the sum of the total predicted time for each pending business.

[0115] In one implementation, a time deviation coefficient is determined based on the target predicted total time and the start processing time. Further, the business processing waiting time for the target user is determined based on the predicted total time, the total number of tellers, and the time deviation coefficient.

[0116] By identifying the services in the "starting" state from all pending services and determining the starting processing time for each service, and using the total predicted processing time for each service as the target total predicted processing time, a time deviation coefficient is determined based on the starting processing time and the target total predicted processing time. Finally, based on the sum of the predicted total processing time, the total number of tellers, and the time deviation coefficient, the waiting time for the target user is determined. The beneficial effects are:

[0117] Firstly, by identifying "start processing status" services, capturing the actual start processing time, comparing it with the predicted total processing time, and calculating the time deviation coefficient, a real-time status tracking mechanism is formed, which can dynamically correct the prediction deviation of service processing waiting time.

[0118] Secondly, by dynamically weighting the time consumption deviation coefficient, the predicted business processing waiting time fluctuates with the current service efficiency, thereby reducing the prediction error rate.

[0119] Thirdly, users can be shown dynamically adjusted service processing times, replacing vague prompts and enhancing user trust.

[0120] Optionally, a time deviation coefficient is determined based on the initial processing time and the target predicted total processing time, including:

[0121] A. If the initial total processing time exceeds the target predicted total processing time, determine the time deviation coefficient based on the ratio between the initial total processing time and the target predicted total processing time.

[0122] The total processing time at the start is the sum of the processing times at the start, and the total predicted processing time at the target is the sum of the total predicted processing time at the target.

[0123] For example, assuming the initial total processing time is current_processing_time and the target predicted total processing time is T_G_current, then when current_processing_time > T_G_current, the time deviation coefficient delta = current_processing_time / T_G_current is determined.

[0124] If the initial processing time exceeds the target predicted total processing time, it indicates a delay in actual business processing. In this case, a time deviation coefficient is determined based on the ratio between the initial processing time and the target predicted total processing time. This coefficient can amplify the estimation of subsequent waiting time to reflect potential delays.

[0125] B. If the initial total processing time is less than or equal to the target predicted total processing time, the time deviation coefficient is determined based on the difference between the initial total processing time and the target predicted total processing time.

[0126] For example, assuming the initial total processing time is current_processing_time and the target predicted total processing time is T_G_current, then when current_processing_time ≤ T_G_current, the time deviation coefficient delta is determined to be 1 - (T_G_current - current_processing_time) * a; where a is the reserved buffer coefficient.

[0127] If the initial total processing time is less than or equal to the target predicted total processing time, it indicates that the actual business processing progress is normal or ahead of schedule. In this case, the time deviation coefficient is determined based on the difference between the initial total processing time and the target predicted total processing time. This can prevent the prediction results from being overly sensitive to short-term efficiency fluctuations and make the prediction results more stable and reasonable.

[0128] Optionally, based on the predicted total processing time, the total number of tellers, and the processing time deviation coefficient, the waiting time for the target user is determined, including:

[0129] The predicted total time is adjusted based on the product of the predicted total time and the time deviation coefficient; the waiting time for the target user is determined based on the ratio between the predicted total time adjustment and the total number of tellers.

[0130] In one implementation, the service processing waiting time is determined as follows:

[0131] T=T_queue*delta / S.active_count

[0132] Where T represents the business processing waiting time, T_queue represents the predicted total time and value, delta represents the time consumption deviation coefficient, and S.active_count represents the total number of tellers.

[0133] By determining the adjustment value for the total predicted time based on the product of the sum of the predicted total time and the time deviation coefficient, and by determining the waiting time for the target user based on the ratio between the adjusted value for the total predicted time and the total number of tellers, the beneficial effects are:

[0134] Firstly, traditional solutions predict service processing wait times based solely on historical averages, ignoring real-time fluctuations. This invention introduces a time-consuming deviation coefficient, which, by multiplying the service processing wait time, more closely reflects actual scenarios.

[0135] Secondly, the adjusted total time is divided by the total number of tellers to quantify the workload of a single teller, thus avoiding prediction failure due to increases or decreases in the number of tellers and improving the accuracy of business processing waiting time prediction.

[0136] Example 3

[0137] Figure 3 This is a schematic diagram of a business processing waiting time prediction device provided in Embodiment 3 of the present invention. It can be applied to predicting the corresponding business processing waiting time for users waiting to process transactions at bank branches. Figure 3 As shown, the device includes:

[0138] The information determination module 31 is used to determine at least one other user whose business processing order is ahead of the target user, and to determine at least one candidate business step included in the pending business of the other user, and to determine the baseline time interval corresponding to the candidate business step.

[0139] The business process time prediction module 32 is used to determine the time correction coefficient based on the other user profiles of the other users and the current teller profile of the current teller, and to correct the baseline time interval based on the time correction coefficient to determine the predicted time of the business process corresponding to the candidate business process; wherein, the current teller is the teller who is handling the pending business.

[0140] The business processing waiting time prediction module 33 is used to determine the total predicted time of the pending business based on the predicted time of the business process, and to determine the business processing waiting time for the target user based on the total predicted time of the business process.

[0141] Optionally, the information determination module 31 is specifically used for:

[0142] Based on the collected voice descriptions of business needs from other users, generate business need description text, determine business need keywords based on the business need description text, and perform semantic role labeling on the business need description text to generate semantic role labeled text.

[0143] Based on the business requirement keywords and the semantic role annotation text, determine at least one candidate business step included in the pending business of the other users.

[0144] Optionally, the information determination module 31 is further used for:

[0145] Based on the business requirement keywords and the semantic role annotation text, generate a description text of the business to be processed;

[0146] The description text of the pending business is input into the target large language model, and the target large language model outputs at least one candidate business step included in the pending business based on the description text of the pending business.

[0147] Optionally, the target large language model is generated in the following manner:

[0148] Based on the sample business requirement description text, determine the sample business requirement keywords and the sample semantic role annotation text;

[0149] Generate sample business description text based on the sample business requirement keywords and the sample semantic role annotation text;

[0150] Obtain the business process tags corresponding to the sample business description text, and fine-tune the pre-trained large language model based on the sample business description text and the business process tags to obtain the target large language model.

[0151] Optionally, the information determination module 31 is further used for:

[0152] Based on the business type of the pending business, the profiles of other users, and the profile of the current teller, generate a business tag to be queried, match the business tag to be queried with the candidate business tags of historical business, and determine at least one target business from the historical business based on the matching result;

[0153] Based on the historical time consumption of the candidate business steps in each of the target business processes, the historical average time consumption and the historical standard deviation time consumption are determined, and the baseline time consumption interval is determined based on the historical average time consumption and the historical standard deviation time consumption.

[0154] Optionally, the business process time prediction module 32 is specifically used for:

[0155] Based on the first historical total time consumption corresponding to each of the historical transactions, determine the global average total time consumption, and based on the second historical total time consumption corresponding to each of the target transactions, determine the target average total time consumption.

[0156] The time correction coefficient is determined based on the ratio between the target total time average and the global total time average.

[0157] Optionally, the business process time prediction module 32 is further used for:

[0158] The lower limit and upper limit of the benchmark time are determined based on the benchmark time interval, and the benchmark time difference is determined based on the difference between the upper limit and the lower limit of the benchmark time.

[0159] The baseline time sum is determined based on the sum between the baseline time lower limit and the baseline time difference, and the predicted time of the business process is determined based on the product of the baseline time sum, the time correction coefficient, and the random number coefficient.

[0160] Optionally, the service processing waiting time prediction module 33 is specifically used for:

[0161] From the pending services, identify the services that are in the "start processing" state and determine the start processing duration for each service.

[0162] The total predicted time for each business transaction is taken as the target total predicted time, and a time deviation coefficient is determined based on the start time and the target total predicted time.

[0163] The business processing waiting time for the target user is determined based on the predicted total time, the total number of tellers, and the time deviation coefficient; wherein, the predicted total time is the sum of the predicted total time for each of the pending businesses.

[0164] Optionally, the service processing waiting time prediction module 33 is further used for:

[0165] If the total processing time exceeds the sum of the target predicted total processing time, the time deviation coefficient is determined based on the ratio between the total processing time and the sum of the target predicted total processing time; wherein, the total processing time is the sum of all the processing times, and the sum of the target predicted total processing time is the sum of all the target predicted total processing times.

[0166] If the total processing time at the start is less than or equal to the target predicted total processing time, the time deviation coefficient is determined based on the difference between the total processing time at the start and the target predicted total processing time.

[0167] Optionally, the service processing waiting time prediction module 33 is further used for:

[0168] The total predicted time adjustment value is determined based on the product of the predicted total time and the time deviation coefficient.

[0169] The business processing waiting time for the target user is determined based on the ratio between the predicted total time adjustment value and the total number of tellers.

[0170] The business processing waiting time prediction device provided in this embodiment of the invention can execute the business processing waiting time prediction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method execution.

[0171] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0172] Example 4

[0173] Figure 4 A schematic diagram of an electronic device 40 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0174] like Figure 4 As shown, the electronic device 40 includes at least one processor 41 and a memory, such as a read-only memory (ROM) 42 or a random access memory (RAM) 43, communicatively connected to the at least one processor 41. The memory stores computer programs executable by the at least one processor. The processor 41 can perform various appropriate actions and processes based on the computer program stored in the ROM 42 or loaded into the RAM 43 from storage unit 48. The RAM 43 may also store various programs and data required for the operation of the electronic device 40. The processor 41, ROM 42, and RAM 43 are interconnected via a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0175] Multiple components in electronic device 40 are connected to I / O interface 45, including: input unit 46, such as keyboard, mouse, etc.; output unit 47, such as various types of monitors, speakers, etc.; storage unit 48, such as disk, optical disk, etc.; and communication unit 49, such as network card, modem, wireless transceiver, etc. Communication unit 49 allows electronic device 40 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0176] Processor 41 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 41 performs the various methods and processes described above, such as methods for predicting business processing wait times.

[0177] In some embodiments, the method for predicting service processing wait times may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 40 via ROM 42 and / or communication unit 49. When the computer program is loaded into RAM 43 and executed by processor 41, one or more steps of the method for predicting service processing wait times described above may be performed. Alternatively, in other embodiments, processor 41 may be configured to perform the method for predicting service processing wait times by any other suitable means (e.g., by means of firmware).

[0178] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transferring data and instructions to the storage system, the at least one input device, and the at least one output device.

[0179] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0180] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0181] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0182] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0183] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0184] 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 this is not limited herein.

[0185] 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 of predicting a service handling waiting time, characterized by, The method includes: Identify at least one other user whose business processing order is ahead of the target user, and identify at least one candidate business step included in the pending business of the other user, and determine the baseline time interval corresponding to the candidate business step; Based on the user profiles of other users and the current teller profile, a time-consuming correction coefficient is determined, and the baseline time-consuming interval is corrected according to the time-consuming correction coefficient to determine the predicted time of the business process corresponding to the candidate business process; wherein, the current teller is the teller who handles the pending business. The total predicted time for the pending business is determined based on the predicted time for each business process, and the waiting time for the target user is determined based on the total predicted time for each business process.

2. The method of claim 1, wherein, The step of determining at least one candidate service stage included in the pending services of the other users includes: Based on the collected voice descriptions of business needs from other users, generate business need description text, determine business need keywords based on the business need description text, and perform semantic role labeling on the business need description text to generate semantic role labeled text. Based on the business requirement keywords and the semantic role annotation text, determine at least one candidate business step included in the pending business of the other users.

3. The method of claim 2, wherein, The step of determining at least one candidate business step included in the pending business of other users based on the business requirement keywords and the semantic role annotation text includes: Based on the business requirement keywords and the semantic role annotation text, generate a description text of the business to be processed; The description text of the pending business is input into the target large language model, and the target large language model outputs at least one candidate business step included in the pending business based on the description text of the pending business.

4. The method of claim 3, wherein, The target large language model is generated in the following manner: Based on the sample business requirement description text, determine the sample business requirement keywords, as well as the sample semantic role annotation text; Generate sample business description text based on the sample business requirement keywords and the sample semantic role annotation text; Obtain the business process tags corresponding to the sample business description text, and fine-tune the pre-trained large language model based on the sample business description text and the business process tags to obtain the target large language model.

5. The method of claim 1, wherein, Determining the baseline time interval corresponding to the candidate business process includes: Based on the business type of the pending business, the profiles of other users, and the profile of the current teller, generate a business tag to be queried, match the business tag to be queried with the candidate business tags of historical business, and determine at least one target business from the historical business based on the matching result; Based on the historical time consumption of the candidate business steps in each of the target business processes, the historical average time consumption and the historical standard deviation time consumption are determined, and the baseline time consumption interval is determined based on the historical average time consumption and the historical standard deviation time consumption.

6. The method of claim 5, wherein, The step of determining the time-consuming correction coefficient based on the user profiles of the other users and the current teller profile of the current teller includes: Based on the first historical total time consumption corresponding to each of the historical transactions, determine the global average total time consumption, and based on the second historical total time consumption corresponding to each of the target transactions, determine the target average total time consumption. The time correction coefficient is determined based on the ratio between the target total average time and the global total average time.

7. The method of claim 1, wherein, The step of correcting the baseline time interval according to the time correction coefficient to determine the predicted time of the business process corresponding to the candidate business process includes: The lower limit and upper limit of the benchmark time are determined based on the benchmark time interval, and the benchmark time difference is determined based on the difference between the upper limit and the lower limit of the benchmark time. The baseline time sum is determined based on the sum between the baseline time lower limit and the baseline time difference, and the predicted time of the business process is determined based on the product of the baseline time sum, the time correction coefficient, and the random number coefficient.

8. The method of claim 1, wherein, The step of determining the service processing waiting time for the target user based on the predicted total service time includes: From the pending services, identify the services that are in the "start processing" state and determine the start processing duration for each service. The total predicted time for each business transaction is taken as the target total predicted time, and a time deviation coefficient is determined based on the start time and the target total predicted time. The business processing waiting time for the target user is determined based on the predicted total time, the total number of tellers, and the time deviation coefficient; wherein, the predicted total time is the sum of the predicted total time for each of the pending businesses.

9. The method of claim 8, wherein, The step of determining the time deviation coefficient based on the start processing time and the target predicted total time includes: If the total processing time exceeds the sum of the target predicted total processing time, the time deviation coefficient is determined based on the ratio between the total processing time and the sum of the target predicted total processing time; wherein, the total processing time is the sum of all the processing times, and the sum of the target predicted total processing time is the sum of all the target predicted total processing times. If the total processing time at the start is less than or equal to the target predicted total processing time, the time deviation coefficient is determined based on the difference between the total processing time at the start and the target predicted total processing time.

10. The method of claim 8, wherein, The step of determining the service processing waiting time for the target user based on the predicted total processing time, the total number of tellers, and the processing time deviation coefficient includes: The total predicted time adjustment value is determined based on the product of the predicted total time and the time deviation coefficient. The business processing waiting time for the target user is determined based on the ratio between the predicted total time adjustment value and the total number of tellers.

11. A device for predicting business processing waiting time, characterized in that, The device includes: The information determination module is used to determine at least one other user whose business processing order is ahead of the target user, and to determine at least one candidate business step included in the pending business of the other user, and to determine the baseline time interval corresponding to the candidate business step; The business process time prediction module is used to determine the time correction coefficient based on the other user profiles of the other users and the current teller profile of the current teller, and to correct the baseline time interval based on the time correction coefficient to determine the predicted time of the business process corresponding to the candidate business process; wherein, the current teller is the teller who handles the pending business. The business processing waiting time prediction module is used to determine the total predicted time of the pending business based on the predicted time of the business process, and to determine the business processing waiting time for the target user based on the total predicted time of the business process.

12. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the business processing waiting time prediction method according to any one of claims 1-10.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a processor to execute the business processing waiting time prediction method according to any one of claims 1-10.

14. A computer program product comprising a computer program that, when executed by a processor, implements a method for predicting business processing waiting time according to any one of claims 1-10.

Images