Work order dynamic allocation method and device based on user emotion analysis, medium and equipment

Abstract

The invention discloses a work order dynamic allocation method and device based on user emotion analysis, a medium and equipment. The invention belongs to the technical field of computers. The method comprises the following steps: acquiring multi-modal information of a work order submitted by a user; the multi-modal information comprises one or more of work order description text information, user voice information and submission behavior information; performing user emotion analysis based on the multi-modal information to obtain an emotion analysis result; acquiring grade information of the user; acquiring a work order processing duration; determining a work order priority according to the emotion analysis result, the grade information and the work order processing duration; and performing engineer allocation on each work order based on the priority of the work order to obtain a dynamic allocation result of engineers processed by each work order. By adopting the technical scheme, the timely work order processing service can be provided for the user based on the user emotion analysis, and the engineer allocation result can be matched with the ability of the engineer, so that the work order processing efficiency of the engineer is improved, and the result of dynamically allocating the engineer is more scientific.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, apparatus, medium, and equipment for dynamically allocating work orders based on user sentiment analysis. Background Art

[0002] With the rapid development of science and technology, a variety of products have emerged. The efficiency of generating and handling work orders for product issues has become a key concern for users and engineers. Regarding work order allocation, traditional systems rely on manual experience or fixed rules, such as direct allocation based on work order type or user level. Engineers are also labeled with single skills, such as simply "network / database," without dynamically evaluating their real-time processing capabilities, such as current load pressure or the efficiency of resolving similar work orders in the past. Furthermore, existing work order allocation methods lack feedback, and the weight formula parameters are fixed, making it impossible to reversely optimize based on actual processing results. Therefore, the current work order allocation and processing methods are unlikely to receive positive reviews from users.

[0003] Therefore, how to accurately assign engineers and dynamically consider various factors, especially users' emotional considerations about the problems that arise, is a technical challenge that needs to be solved urgently in this field. Summary of the Invention

[0004] This application proposes a method, device, medium and equipment for dynamic work order allocation based on user sentiment analysis to solve the problems of traditional work order allocation and processing, poor work order processing timeliness, and failure to consider user emotional factors. The technical solution provided by this application proposes a method for dynamic work order allocation based on user sentiment analysis. This method can prioritize work orders based on user sentiment analysis results, level information, and work order processing time, and dynamically match engineers one by one to provide users with timely work order processing services. The engineer allocation results can match the engineer's capabilities, improve the engineer's work order processing efficiency, and the results of dynamic engineer allocation are more scientific.

[0005] This embodiment of the present application provides a method for dynamically allocating work orders based on user sentiment analysis, the method comprising:

[0006] Obtaining multimodal information of a work order submitted by a user; wherein the multimodal information includes one or more of work order description text information, user voice information, and submission behavior information;

[0007] Performing user sentiment analysis based on the multimodal information to obtain a sentiment analysis result;

[0008] Get the user's level information;

[0009] Get the work order processing time;

[0010] Determine the work order priority according to the emotion analysis result, the level information, and the work order processing duration;

[0011] Based on the work order priority, allocate engineers to each work order in sequence to obtain the dynamic allocation result of the engineers handling each work order.

[0012] Further, based on the work order priority, allocate engineers to each work order in sequence to obtain the dynamic allocation result of the engineers handling each work order, including:

[0013] Based on the work order priority in sequence, determine the work orders to be allocated in the order from high to low priority;

[0014] Obtain the static skill information of the engineers and the dynamic matching information for the work orders to be allocated;

[0015] Generate a list of engineer matching degrees for the work orders to be allocated according to the static skill information and the dynamic matching information;

[0016] Based on the engineer sorting in the list of engineer matching degrees, determine the target engineer for the work order to be allocated, and traverse all work orders to obtain the dynamic allocation result of the engineers handling each work order.

[0017] Further, determine the work order priority according to the emotion analysis result, the level information, and the work order processing duration, including:

[0018] Obtain the emotion score corresponding to the emotion analysis result, the level score corresponding to the level information, and the duration score corresponding to the work order processing duration;

[0019] Obtain the emotion weight value, the level weight value, and the processing duration weight value;

[0020] Determine the priority score of the work order according to the product of the emotion score and the emotion weight value, the product of the level score and the level weight value, and the product of the duration score and the processing duration weight value;

[0021] Determine the work order priority according to the priority scores of all work orders.

[0022] Further, the emotion score is determined based on the interval in which the emotion analysis result falls;

[0023] The level score is determined based on the interval in which the level information falls;

[0024] The duration score is determined based on the interval in which the work order processing duration falls.

[0025] Furthermore, user sentiment analysis is performed based on the multimodal information to obtain sentiment analysis results, including:

[0026] User sentiment analysis is performed based on the number of keywords and punctuation features of the work order description text information, the speaking speed features, pitch features, and pause frequency features of the user voice information, as well as the submission frequency features and interaction path features of the submission behavior in the multimodal information to obtain a sentiment analysis result.

[0027] Furthermore, the method further comprises:

[0028] When a newly generated work order is identified and the work order priority of the newly generated work order exceeds the set threshold, the work order to be assigned is re-determined and the engineer is dynamically assigned again.

[0029] Furthermore, after obtaining the dynamic allocation results of the engineers processing each work order, the method further includes:

[0030] Obtain the actual processing time of each work order, engineer feedback, and user feedback to adjust the considerations for engineer allocation.

[0031] The embodiment of the present application further provides a device for dynamically allocating work orders based on user sentiment analysis, the device being configured on a backend server and comprising:

[0032] A multimodal information acquisition module, configured to acquire multimodal information of a work order submitted by a user; wherein the multimodal information includes one or more of work order description text information, user voice information, and submission behavior information;

[0033] A sentiment analysis result determination module, configured to perform user sentiment analysis based on the multimodal information to obtain a sentiment analysis result;

[0034] Level information acquisition module, used to obtain the user's level information;

[0035] The work order processing time acquisition module is used to obtain the work order processing time;

[0036] A work order priority determination module, configured to determine the work order priority based on the sentiment analysis result, the level information, and the work order processing time;

[0037] The dynamic allocation module is used to allocate engineers to each work order in order based on the work order priority, and obtain the dynamic allocation result of the engineer processing each work order.

[0038] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method for dynamic work order allocation based on user sentiment analysis.

[0039] An embodiment of the present application further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the method for dynamically allocating work orders based on user emotion analysis as described above is implemented.

[0040] The embodiment of the present application adopts the following technical solutions: obtaining multi-modal information of a work order submitted by a user; wherein, the multi-modal information includes one or more of work order description text information, user voice information, and submission behavior information; performing user emotion analysis based on the multi-modal information to obtain an emotion analysis result; obtaining the level information of the user; obtaining the work order processing duration; determining the work order priority according to the emotion analysis result, the level information, and the work order processing duration; and dynamically allocating engineers to each work order in sequence based on the work order priority to obtain a dynamic allocation result of the engineer for each work order processing.

[0041] At least one of the above technical solutions adopted by the embodiment of the present application can achieve the following beneficial effects:

[0042] This solution proposes a method for dynamically allocating work orders based on user emotion analysis. This method can sort the priorities of work orders according to the user emotion analysis result, level information, work order processing duration, etc., and dynamically match engineers one by one, providing timely work order processing services for users. Moreover, the engineer allocation result can be matched with the engineer's ability, improving the work order processing efficiency of engineers, and the result of dynamically allocating engineers is more scientific. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation to the present application. In the drawings:

[0044] Figure 1 is a schematic flow chart of the method for dynamically allocating work orders based on user emotion analysis provided in Embodiment 1 of the present application;

[0045] Figure 2 is a schematic structural diagram of the system for dynamically allocating work orders based on user emotion analysis provided in Embodiment 2 of the present application;

[0046] Figure 3 is a schematic structural diagram of the device for dynamically allocating work orders based on user emotion analysis provided in Embodiment 3 of the present application;

[0047] Figure 4 is a schematic structural diagram of an electronic device provided in Embodiment 4 of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments of this application and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts shall fall within the scope of protection of this application.

[0049] The following will detail the technical solutions provided by each embodiment of this application in conjunction with the drawings.

[0050] Embodiment 1

[0051] Figure 1 It is a flowchart of the work order dynamic allocation method based on user emotion analysis provided for Embodiment 1 of this application. As Figure 1 shown, the method includes:

[0052] S11, obtaining multi-modal information of the work order submitted by the user; wherein, the multi-modal information includes one or more of work order description text information, user voice information, and submission behavior information;

[0053] Among them, the user refers to an individual or organization that uses the system and submits a work order, which can be an enterprise customer, an individual consumer, etc. In this scenario, it is the initiator of the work order and can provide various relevant information.

[0054] A work order is a task request submitted by the user that needs to be processed by the system or relevant personnel, and it contains specific problems or requirements, such as equipment failure repair requests, service consultations, etc.

[0055] Multi-modal information can refer to information containing various different forms. In this scenario, it covers work order description text information, which is a description of the work order problem, requirements, etc. in text form, user voice information, which is the content of the work order expressed by the user through voice, such as a detailed description of the problem, etc., and submission behavior information, which is information related to the operation behavior when the user submits the work order, such as submission time, submission location, etc.

[0056] Work order description text information can be that the user elaborates on the specific content of the work order in detail in text, such as the fault phenomenon, the problem to be solved, etc.

[0057] User voice information can be that the user conveys the information about the work order to the system through voice input, and the system can convert it into processable text data through voice recognition technology.

[0058] Submission behavior information can be the relevant data of the user's submission behavior of the work order, such as the submission frequency, whether it is submitted concentratedly in a specific time period, etc.

[0059] This solution can obtain relevant multimodal information from the user side through specific technical means, such as receiving through network interfaces and collecting through sensors. For example, for text information, it can be read by the system after the user enters it in a web form; for voice information, it can be collected using a microphone and converted into text through speech recognition; for submission behavior information, it can be obtained by the system recording the timestamp, IP address, etc. of the user's operations.

[0060] S12, perform user emotion analysis based on the multimodal information to obtain an emotion analysis result;

[0061] User emotion can be the emotional tendency shown by the user when submitting a work order, such as anger, anxiety, calmness, satisfaction, etc.

[0062] The emotion analysis result can be a specific conclusion about the user's emotion obtained through the processing and analysis of multimodal information, such as the category of emotion (positive, negative, neutral) and the intensity level of the emotion, etc.

[0063] This solution can process and analyze multimodal information. For example, natural language processing technology can be used to analyze the emotional words in text information, and speech intonation analysis technology can be used to process the emotional characteristics in voice information. By synthesizing these analyses to judge the user's emotion. Through the above emotion analysis operation, the final emotion analysis result is output, and this result can be presented in the form of data or labels for subsequent processing.

[0064] S13, obtain the user's level information;

[0065] The level information can be an identifier used to measure the importance or value of the user in the system. For example, membership levels, such as ordinary members, senior members, and VIP members, credit levels, etc. Different levels may correspond to different service priorities or rights and interests.

[0066] This solution can read the user's level information from the system's user database or relevant storage, such as obtaining the corresponding information by querying the level field recorded in the user table.

[0067] S14, obtain the work order processing duration;

[0068] The work order processing duration can reflect the urgency of the work order to be processed. For example, if the work order processing duration is 30 minutes, it is a relatively urgent work order, and for some less urgent work orders, the work order processing duration can be 12 hours or 24 hours.

[0069] This solution can determine the work order processing duration of the current work order by identifying the work order type, etc.

[0070] S15. Determine the work order priority according to the emotion analysis result, the level information, and the work order processing duration.

[0071] This solution can use the emotion analysis result, level information, and work order processing duration obtained previously as the basis for determining the work order priority. For example, by using a specific algorithm or rule, comprehensively analyze and calculate the above data to obtain the priority value or level of each work order. For example, weights can be set for different factors, and then the priority score can be obtained through weighted calculation.

[0072] S16. Based on the work order priority, allocate engineers to each work order in sequence to obtain the dynamic allocation result of the engineers for each work order.

[0073] An engineer is a professional responsible for processing work orders, with different skills and experiences, and is reasonably allocated according to the type and difficulty of the work order.

[0074] The dynamic allocation result can be the result obtained by allocating work orders based on the work order priority and the real-time status of engineers, such as busyness and skill matching, which reflects which engineer is responsible for processing each work order.

[0075] This solution can be based on the work order priority and operate in order from high to low according to the work order priority. Specifically, the allocation of engineers can be carried out one by one, then the work order information can be sent to the corresponding engineer, and the allocation status of the work order in the system can be updated. Through the above engineer allocation operation, the dynamic allocation result is output, which can be recorded in the system database for query and management.

[0076] This technical solution can comprehensively understand the user's needs and emotional state by obtaining the multi-modal information of the work order submitted by the user. Combining the user level information and the work order processing duration, it can scientifically and reasonably determine the work order priority, so that important and urgent work orders can be processed first. Based on the work order priority, the dynamic allocation of engineers is carried out, which improves the efficiency and accuracy of work order processing, optimizes resource allocation, and enhances user satisfaction. It has significant application value especially for work order management in the customer service scenario, and can better meet user needs and improve service quality.

[0077] The technical solution provided in this embodiment obtains the multimodal information of the work orders submitted by users; wherein, the multimodal information includes one or more of the work order description text information, user voice information, and submission behavior information; performs user emotion analysis based on the multimodal information to obtain an emotion analysis result; obtains the level information of the user; obtains the work order processing duration; determines the work order priority according to the emotion analysis result, the level information, and the work order processing duration; and performs engineer allocation on each work order in sequence based on the work order priority to obtain the dynamic allocation result of the engineer for each work order processing. This solution can sort the work orders according to the user emotion analysis result, level information, work order processing duration, etc., and dynamically match engineers one by one, providing timely work order processing services for users, and the engineer allocation result can be matched with the engineer's ability, improving the work order processing efficiency of the engineer, and the result of dynamically allocating engineers is more scientific.

[0078] In one embodiment, optionally, performing engineer allocation on each work order in sequence based on the work order priority to obtain the dynamic allocation result of the engineer for each work order processing includes:

[0079] Determine the work orders to be allocated in sequence based on the work order priority, in the order from high to low priority.

[0080] Obtain the static skill information of the engineer and the dynamic matching information for the work orders to be allocated.

[0081] Generate a list of engineer matching degrees for the work orders to be allocated according to the static skill information and the dynamic matching information.

[0082] Determine the target engineer for the work orders to be allocated based on the engineer sorting in the list of engineer matching degrees, and traverse all work orders to obtain the dynamic allocation result of the engineer for each work order processing.

[0083] Among them, the work orders to be allocated can refer to the work orders that have not been allocated to engineers for processing after being sorted from high to low according to the work order priority. These work orders wait for the system to find suitable engineers to be responsible for processing according to certain rules.

[0084] The static skill information can be the relatively fixed skill-related information possessed by the engineer. For example, the professional technical qualification certificates possessed by the engineer, such as network engineer certificates, software engineer certificates, etc., the technical fields they are good at, such as database management, hardware repair, etc., and the programming languages they master. These information do not change with the real-time changes of the work order processing situation and are the basic ability attributes of the engineer himself.

[0085] Dynamic matching information can be real-time changing matching-related information related to engineers for work orders to be assigned. For example, the current workload of an engineer, the number of work orders being processed, the estimated remaining processing time, etc., and the relevance between the work order to be assigned and the engineer's current work, such as whether the work orders being processed and the work order to be assigned belong to the same technical field. This information is updated in real time as the work orders are assigned and processed, reflecting the current actual work status and matching possibilities.

[0086] The engineer matching degree list can be a list generated by quantitatively evaluating the matching degree between each work order to be assigned and all engineers based on the static skill information of the engineers and the dynamic matching information for the work orders to be assigned. The list usually contains the identifiers of the engineers and the corresponding matching degree scores or rankings. The engineers with higher matching degrees are ranked higher in the list.

[0087] The target engineer can be the engineer who is most suitable for processing a specific work order to be assigned, determined according to the ranking in the engineer matching degree list for each work order to be assigned. That is, the engineer with the highest matching degree is selected as the target engineer to process the corresponding work order.

[0088] This solution can be based on the work order priority and be assigned in the order from high to low work order priority. By sorting the work order priorities, the work orders in the to-be-assigned state are screened out and marked as work orders to be assigned for subsequent engineer assignment operations. Read the static skill information of the engineers from the relevant database or information storage of the system, and obtain the dynamic matching information about the work orders to be assigned in real time. Obtaining the static skill information may involve querying the engineer information table, while obtaining the dynamic matching information requires real-time monitoring and calculation of data such as the current workload. According to the obtained static skill information of the engineers and the dynamic matching information of the work orders to be assigned, use specific algorithms or models to calculate and evaluate the matching degree between each engineer and the work order to be assigned, thereby generating the engineer matching degree list. Based on the engineer matching degree list, obtain the ranking information of the engineers from the list to determine the target engineer. Specifically, according to the ranking in the engineer matching degree list, select the most suitable engineer for each work order to be assigned, that is, the target engineer, to clarify the person responsible for processing the work order. Perform the above operations of determining the target engineer for all work orders in sequence to ensure that each work order can find a suitable engineer for processing, thus obtaining the complete dynamic assignment result of the engineers for each work order.

[0089] In this technical solution, by determining the work orders to be assigned in an orderly manner based on the work order priority, and comprehensively considering the static skill information of engineers and the dynamic matching information for the work orders, a list of engineer matching degrees is generated to determine the target engineer, achieving a scientific and reasonable matching of work orders and engineers. This dynamic allocation method can make full use of the professional skills of engineers, give priority to handling high-priority work orders, and at the same time consider the real-time workload of engineers to avoid uneven distribution of engineer tasks. By traversing all work orders, it ensures that each work order can be properly processed, improves the efficiency and quality of work order processing, optimizes resource allocation, and enhances the overall service response speed and user satisfaction.

[0090] In one embodiment, optionally, determining the work order priority according to the emotion analysis result, the level information, and the work order processing duration includes:

[0091] Obtain the emotion score corresponding to the emotion analysis result, the level score corresponding to the level information, and the duration score corresponding to the work order processing duration;

[0092] Obtain the emotion weight value, the level weight value, and the processing duration weight value;

[0093] Determine the priority score of the work order according to the product of the emotion score and the emotion weight value, the product of the level score and the level weight value, and the product of the duration score and the processing duration weight value;

[0094] Determine the work order priority according to the priority scores of all work orders.

[0095] Among them, the emotion score can be a numerical value obtained by quantifying the emotion analysis result. Different emotional states, such as anger, anxiety, and calmness, will correspond to different scores. For example, anger may correspond to a relatively high negative score, while calmness corresponds to a relatively low fluctuation score, so as to intuitively reflect the intensity and tendency of the user's emotion.

[0096] The level score can be a quantified numerical value corresponding to the user level information. Different user levels, such as ordinary members, premium members, and VIP members, are assigned different scores. Higher levels usually correspond to higher scores, reflecting the differences in the importance of users in the system.

[0097] The duration score can be a quantified numerical value corresponding to the work order processing duration. Different scores can be divided according to a preset duration interval. For example, the longer the processing duration, the higher the corresponding score, to reflect the urgency or complexity of the work order processing.

[0098] The emotional weight value can be a coefficient used to measure the importance of the emotional score in determining the work order priority. The larger the weight value, the greater the impact of the emotional factor on the work order priority.

[0099] The level weight value can be a coefficient reflecting the importance of the level score in the process of determining the work order priority. A high level weight value means that the user level has a more significant impact on the work order priority.

[0100] The processing duration weight value can be a coefficient indicating the importance of the duration score in determining the work order priority. A high weight value indicates that the work order processing duration has a greater impact on the priority.

[0101] The priority score can be a value obtained by multiplying the emotional score, the level score, and the duration score by their respective weight values and then adding them together, comprehensively reflecting the comprehensive importance and processing priority of a work order.

[0102] The work order priority can be the processing sequence determined by sorting the work orders according to the priority scores of all work orders. Work orders with higher priorities should be processed first.

[0103] Specifically, when calculating the priority score, according to the formula:

[0104] Priority score = Emotional score × Emotional weight value + Level score × Level weight value + Duration score × Processing duration weight value;

[0105] Thus, the priority score of each work order is determined.

[0106] According to the priority scores of all work orders, the work orders are sorted, and then the priority order of each work order is determined. For example, the priority scores can be sorted from high to low, and the work orders with higher scores have higher priorities.

[0107] This solution realizes the objective and scientific evaluation of the work order priority through the method of quantifying and weighted calculation. It comprehensively considers multiple key factors such as user emotions, user levels, and work order processing durations, avoiding unreasonable priority allocations that may be dominated by a single factor. By reasonably setting the weight values, the influence of each factor can be flexibly adjusted according to different business scenarios and requirements, improving the pertinence and efficiency of work order processing. High-priority work orders can be processed in a timely manner, which helps to improve user satisfaction. In particular, work orders from users with negative emotions or high levels can be given priority responses, and at the same time, work orders with longer processing durations can also be taken into account, optimizing the overall work order processing process and resource allocation.

[0108] In one embodiment, optionally, the emotional score is determined based on the interval in which the emotional analysis result falls;

[0109] The level score is determined based on the interval in which the level information falls;

[0110] The duration score is determined based on the interval in which the work order processing duration falls.

[0111] The interval can refer to several pre-set range segments for quantifying different types of information, such as emotions, user levels, work order processing durations, etc. Each interval corresponds to a specific score range or specific score. For example, for the result of sentiment analysis, it may be set that the interval corresponding to the "angry" emotion is 80 - 100 points, and the interval corresponding to the "calm" emotion is 40 - 60 points, etc.

[0112] The level score can be a numerical value after quantifying the user level information, reflecting the importance of the user level in determining the work order priority, and the specific value is determined according to the corresponding interval in which the level information falls.

[0113] The duration score can be a numerical value after quantifying the work order processing duration, used to reflect the influence degree of the work order processing duration factor when determining the work order priority, and the specific value is determined according to the specific interval in which the work order processing duration falls.

[0114] In this technical solution, the method of determining various scores based on intervals quantifies and standardizes the originally relatively vague information (such as user emotions, user levels, work order processing durations), enabling more accurate and objective measurement of the influence of each factor when determining the work order priority. By pre-setting intervals and corresponding scores, a unified standard is established, avoiding the uncertainty of subjective judgment. At the same time, this method has a certain degree of flexibility and can adjust the interval range and corresponding scores according to actual business requirements and scenarios, so as to better adapt to different situations. During the process of determining the work order priority, these quantified scores can be more conveniently involved in the calculation, improving the scientificity and rationality of priority determination, helping to optimize the work order processing process, ensuring that resources can be more reasonably allocated to work orders of different priorities, and ultimately enhancing the overall service quality and user satisfaction.

[0115] In one embodiment, optionally, user sentiment analysis is performed based on the multimodal information to obtain a sentiment analysis result, including:

[0116] User sentiment analysis is performed according to the number of keywords and punctuation mark features in the work order description text information, the speech rate feature, pitch feature, and pause frequency feature of the user voice information, as well as the submission frequency feature and interaction path feature of the submission behavior in the multimodal information to obtain a sentiment analysis result.

[0117] Multimodal information can contain a variety of different forms of information related to the work order submitted by the user, including work order description text information, user voice information, and submission behavior information. These different forms of information can reflect the user's status when submitting the work order from multiple angles.

[0118] The work order description text information can be a detailed description of the work order problem, requirements, etc. in text form by the user. It is an important carrier for users to express their intentions and emotions.

[0119] The number of keywords can be the number of words with specific semantics and sentiment in the ticket description text. For example, a high number of negative keywords such as "serious" and "unbearable" may indicate that the user has a negative sentiment.

[0120] Punctuation features can be the characteristics reflected by the punctuation marks used in the work order description text. For example, the extensive use of exclamation marks may indicate that the user is excited or angry; the frequent use of ellipsis marks may reflect the user's helplessness and other emotions.

[0121] User voice information can be relevant content about the work order entered by the user through voice. Various voice features can reflect the user's emotional state when speaking.

[0122] The speaking speed feature can be the speaking speed of the user's voice. A fast speaking speed may indicate that the user is nervous or excited; a slow speaking speed may indicate that the user is depressed or in a thinking state.

[0123] Tone features can be the pitch changes of voice. High pitch may reflect the user's strong emotions such as excitement and anger; low pitch may indicate that the user is calm or depressed.

[0124] The pause frequency feature can be the frequency of pauses in the user's speech. Frequent pauses may indicate that the user is hesitant, nervous, or emotional when expressing, resulting in unfluent speech.

[0125] Submission behavior information can be relevant data reflecting the process of users submitting work orders, which can indirectly reflect the user's emotions and the urgency of their needs.

[0126] The submission frequency feature can be the frequency with which users submit work tickets. Submitting work tickets multiple times in a short period of time may indicate that the user is eager to solve the problem and is relatively anxious.

[0127] The interaction path features can be the operation steps and page paths that users go through when submitting a work order. Complex or abnormal interaction paths may indicate that users encounter difficulties during the operation, thereby generating negative emotions.

[0128] The result of sentiment analysis can be a conclusion about the user's emotional state obtained after analyzing various types of features mentioned above, such as specific emotional categories like positive, negative, neutral, etc., and the intensity level of the emotion.

[0129] This technical solution, a method for user sentiment analysis based on various features in multimodal information, makes full use of various forms of data and can capture the user's emotional state more comprehensively and accurately when submitting a work order. It analyzes from multiple dimensions including text, voice, and submission behavior, avoiding the one-sidedness and inaccuracy that may be brought by a single information source. For example, text information may not be as true due to the user deliberately hiding emotions, while voice and submission behavior features can be used as supplements for verification and correction. Accurate sentiment analysis results help to more reasonably determine the work order priority, enabling the preferential processing of user work orders with negative emotions, improving user satisfaction, and also helping the enterprise to better understand user needs and experiences, and optimize service processes and product quality.

[0130] In one embodiment, optionally, the method further includes:

[0131] When a newly generated work order is recognized and the work order priority of the newly generated work order exceeds a set threshold, re-determine the work order to be assigned and re-perform the dynamic allocation of engineers.

[0132] A newly generated work order can be a work order newly created in the system that has not been assigned for processing yet. It may be a work order request newly submitted by the user or a work order automatically generated by the system according to certain rules.

[0133] The set threshold can be a specific priority value or level standard pre-set in the system. When the priority of the newly generated work order exceeds this threshold, corresponding operations will be triggered. This threshold can be adjusted according to business requirements and actual situations. For example, set the threshold for high-priority work orders to ensure that important work orders can be processed in a timely manner.

[0134] In this solution, the system detects and determines whether there are newly generated work orders through specific algorithms, rules, or monitoring mechanisms, and obtains relevant information about the work orders, including their priorities, etc. For example, the identification operation can be achieved by listening for work order creation events or periodically querying the work order database. Compare the priority value or level of the newly generated work order with the set threshold. When the value of the work order priority is greater than the value of the set threshold, or the level of the work order priority is higher than the level of the set threshold, it is considered to "exceed" the set threshold. When the priority of the newly generated work order exceeds the set threshold, the current set of work orders to be assigned is re-evaluated and screened. It is possible to include the newly generated high-priority work orders in the set of work orders to be assigned, and at the same time adjust the order and status of the original work orders to be assigned according to the new situation. After re-determining the work orders to be assigned, the operation process of dynamically assigning engineers is performed again according to factors such as work order priority and the skills and workload of engineers, so as to ensure that high-priority work orders can be assigned to appropriate engineers for processing as soon as possible.

[0135] The mechanism of this solution to re-determine the work orders to be assigned and re-perform the dynamic assignment of engineers when the priority of the newly generated work order exceeds the set threshold can respond in a timely manner to the emergence of high-priority work orders, ensuring that important work orders will not be delayed in processing. By flexibly adjusting the set of work orders to be assigned and re-assigning engineers, the adaptability and responsiveness of the work order processing system are improved, enabling resources to be more reasonably allocated to the work orders that most need them. It avoids the backlog of high-priority work orders in the waiting queue, improves the overall efficiency and service quality of work order processing, thereby enhancing users' satisfaction and trust in the system, and has important practical significance for business scenarios that require timely handling of important affairs.

[0136] In one embodiment, optionally, after obtaining the dynamic assignment results of the engineers for each work order processing, the method further includes:

[0137] Obtain the actual processing duration, engineer feedback, and user feedback of each work order, and adjust the consideration factors for engineer assignment.

[0138] The actual processing duration can be the time length actually spent from the start of processing to the final completion of each work order. It reflects the efficiency of engineers in processing work orders and may differ from the previously estimated work order processing duration.

[0139] Engineer feedback can be information provided by engineers to the system or relevant management personnel during or after processing work orders, regarding aspects such as the processing situation of the work order, problems encountered, and required resources. For example, an engineer may feedback that a certain work order is difficult and cannot be efficiently solved with existing skills, or that the information of a certain work order is incomplete, etc.

[0140] User feedback can be the evaluation and opinions given by users on aspects such as the processing result, service quality, and processing duration after the work order is processed. User feedback can help understand the user's satisfaction with the work order processing and whether the expectations are met.

[0141] The consideration factors for engineer assignment can be various factors based on which dynamic engineer assignment of work orders is carried out, such as work order priority, static skill information of engineers, dynamic matching information, user sentiment analysis results, and user level information, etc. These factors together determine which engineer is most suitable for each work order.

[0142] This solution can modify and optimize the consideration factors for engineer assignment according to the actual processing duration of each work order, engineer feedback, and user feedback information obtained. For example, if it is found that a certain engineer has an overly long actual processing duration when processing a specific type of work order, it may be necessary to adjust the skill matching weight of this type of work order and this engineer; if the user feedback is dissatisfied with a certain engineer's service, it may be necessary to consider reducing the priority of this engineer in subsequent work order assignments, etc.

[0143] This technical solution can make the work order assignment mechanism more perfect and reasonable by obtaining the actual processing duration of each work order, engineer feedback, and user feedback, and adjusting the consideration factors for engineer assignment. The actual processing duration can help evaluate the work efficiency and ability of engineers and provide a more accurate reference for subsequent assignments; engineer feedback can timely discover problems existing in the work order processing process and optimize the assignment strategy accordingly; user feedback directly reflects the service quality and user satisfaction, prompting the system to continuously improve. This continuously optimized mechanism can improve the efficiency and quality of work order processing, better meet user needs, enhance user satisfaction, and at the same time help to reasonably allocate engineer resources, improve the work enthusiasm and work effect of engineers, and achieve the virtuous cycle and continuous development of the work order processing system.

[0144] Embodiment 2

[0145] Figure 2 It is the system structure diagram of the dynamic assignment of work orders based on user sentiment analysis provided by Embodiment 2 of this application. As Figure 2 shown, the system architecture of this solution is as follows:

[0146] 1. Multimodal data acquisition module:

[0147] Text analysis: Extract keywords in the work order description (such as "crash", "urgent"), punctuation features (exclamation mark density);

[0148] Voice analysis: Analyze the speech rate, pitch, and pause frequency of the user's voice work order, and quantify the anxiety index (such as a speech rate > 200 words per minute is counted as 0.8);

[0149] Behavior analysis: Record the user's submission frequency (e.g., repeated submissions within 5 minutes are counted as "urgent") and interaction path (e.g., multiple clicks on the "expedite" button).

[0150] 2. Emotional grading and quantification engine:

[0151] User emotion index (highest is 1):

[0152] Level 1 (calm): The text has no negative words and the voice is stable (score 0.1 - 0.3);

[0153] Level 2 (anxious): The text contains 0 - 1 negative words and the voice is slightly higher <80db> 55db (score 0.4 - 0.6);

[0154] Level 3 (angry): The text contains more than 3 negative words, the voice pitch > 80dB, and the speech rate increases (score 0.7 - 1.0).

[0155] 3. User value analysis engine:

[0156] User value index (highest is 1):

[0157] VIP level: Distinguish the index according to the VIP level (score 0.3 - 0.7, 0 if there is no VIP);

[0158] Historical behavior: Based on the user's historical behavior, such as the complaint rate being greater than 30% (score 0 - 0.3).

[0159] 4. SLA (work order processing duration) weighting engine:

[0160] SLA weighting index (highest is 1):

[0161] If the remaining SLA time is short, the weighting is high: If there is only half an hour left to complete the work order, the weighting is 1;

[0162] If the remaining SLA time is long, the weighting is low: If there are 24 hours left to process the work order, the weighting is 0.5.

[0163] 5. Engineer skill engine:

[0164] Engineer skill index (highest is 1): Static skill index + Dynamic skill index

[0165] Static skill index: Whether the engineer is an expert in this type of problem (such as network, database, cloud platform);

[0166] Dynamic skill index: Real-time ability assessment based on historical work order processing data (such as the success rate and efficiency of an engineer in processing similar work orders).

[0167] 6. Dynamic Weight Calculation Engine:

[0168] Dynamic formula:

[0169] Priority = (SLA Weighting Index × W1) + (Engineer Skill Index × W2) + (User Emotion Index × W3) + (User Value Index × W4);

[0170] Among them, the SLA weighting weight W1 is adjusted according to whether it is the peak period, with a maximum of 0.4; the engineer skill weight W2 is adjusted according to the engineer skill index, with a maximum of 0.5; the user emotion index weight W3 is adjusted according to the user emotion index, with a maximum of 0.5; the user value index weight W4 is adjusted according to the user value index, with a maximum of 0.6;

[0171] And, it can be specified that W1 + W2 + W3 + W4 ≤ 2.

[0172] Engineer Portrait (generate a pending queue according to engineers):

[0173] Real-time load pressure = current number of work orders / daily processing capacity;

[0174] Historical processing capacity = historical resolution rate of similar work orders × proportion of urgent work orders.

[0175] 7. Global Optimization and Allocation Module:

[0176] Allocation mode: Allocate work orders according to the engineer weight queue and the engineer pending queue.

[0177] Allocation algorithm: Improved Hungarian algorithm + preemptive reallocation mechanism; Assume that engineer A has a high weight, but due to reasons such as load pressure, other more suitable engineer B will be assigned to handle the work order.

[0178] Preemption condition: Trigger reallocation when the weight of the new work order > 20% of the weight of the assigned but unprocessed work order. Or when the SLA has triggered the threshold and urgent allocation is required, high-skilled personnel will be given priority to handle.

[0179] 8. Feedback and Learning Module:

[0180] Parameter optimization: Update the weight coefficients based on the work order processing results, such as resolution duration, user rating, and engineer feedback, etc.:

[0181] The processing flow of this system can include the following steps:

[0182] 1. Work Order Submission and Multimodal Analysis:

[0183] The user submits a work order (text / voice) from the terminal;

[0184] The system analyzes the emotion level and calculates the user emotion index;

[0185] The system calculates the SLA weighted index based on the SLA weighted sum and the peak period.

[0186] The system obtains information such as the user's VIP level and calculates the user value index.

[0187] The system obtains the list of on-duty engineers and calculates the engineer skill level.

[0188] 2. Generate a dynamic weight queue according to the engineer list:

[0189] Example: A VIP3 user has a historical complaint rate of 0, an emotion classification of Level2, an engineer A skill index of 0.7, 3 hours remaining for the SLA, and it is the peak period.

[0190] The dynamic weight values are: W1 = 0.4, W2 = 0.4, W3 = 0.3, W4 = 0.5;

[0191] Priority = (SLA weighted index × W1) + (engineer skill index × W2) + (user emotion index × W3) + (user value index × W4);

[0192] For example, the priority of engineer A is: Priority(engineer A) = (0.5 × W1) + (0.7 × W2) + (0.6 × W3) + (0.5 × W4) = 0.91;

[0193] The full score of the priority is 2. In case of extremely urgent situations, a temporary additional value can be specified as a special handling solution for increasing the weight. For example, for all work orders during the Double Eleven promotion period, an additional weight of 1 is added for emergency avoidance;

[0194] 3. Generate an off-duty queue according to the engineer list:

[0195] Generate the engineer off-duty queue based on the real-time load pressure and historical completion rate;

[0196] Real-time load pressure = current number of work orders / daily processing capacity;

[0197] Historical completion rate = historical resolution rate of similar work orders × proportion of emergency work orders.

[0198] 4. Work order allocation and preemption:

[0199] Allocation: Optimally allocate work orders according to the engineer weight queue and the engineer off-duty queue;

[0200] Weight preemption: If the weight of a new work order > 20% of the weight of the already allocated work order, withdraw the work order with a low weight and re-allocate it.

[0201] SLA Preemption: If the remaining time of the SLA has triggered the alarm threshold, high-matching engineers will be forcibly assigned to handle the issue first, and the remaining work orders of these engineers will be postponed or reassigned.

[0202] 5. Feedback and Learning:

[0203] Work order data feedback: Work order duration, engineer feedback, user scoring, high-emotion user satisfaction, VIP user satisfaction, engineer assignment plan satisfaction, and adjust weights based on multiple data sources.

[0204] In an exemplary scenario, such as in the complaint handling of the financial industry, the following process may be included:

[0205] Identify the anxiety emotion (Level 3) in the voice of high-level VIP customers, dynamically increase the weight W4 to 0.6, and due to the high comprehensive weight score, assign it to a senior assistant or customer service manager for handling first, and solve the problem within 15 minutes;

[0206] During the peak sales period of e-commerce:

[0207] Automatically increase the SLA weighted weight to 0.4 to ensure that promotion-related work orders are preferentially assigned.

[0208] This technical solution can perform multi-modal emotion quantification, integrate text, voice, and behavior data, construct a hierarchical emotion model, and break through the limitations of single text analysis. Moreover, it can adaptively adjust the weight coefficients, such as dynamically adjusting the formula coefficients according to business scenarios (peak / daily), user value, and processing feedback. It also refines the engineer profile, introduces real-time load pressure and historical emergency work order resolution rate, and optimizes the calculation of skill matching degree. In addition, this solution can also output an interpretable assignment report and generate a work order assignment basis report, such as "due to the user's emotion Level 3 and high VIP level, the weight is increased to a high value, and combined with matching an engineer with the corresponding skills, the problem is solved in a short time and receives good reviews".

[0209] Embodiment 3

[0210] Figure 3 It is a schematic structural diagram of a work order dynamic assignment device based on user emotion analysis provided by Embodiment 3 of this application. The device is configured in a background server, as Figure 3 shown, the device includes:

[0211] A multi-modal information acquisition module 310, which is used to acquire multi-modal information of the work order submitted by the user; wherein, the multi-modal information includes one or more of work order description text information, user voice information, and submission behavior information;

[0212] An emotion analysis result determination module 320, which is used to perform user emotion analysis based on the multi-modal information to obtain an emotion analysis result;

[0213] A level information acquisition module 330, configured to acquire the level information of a user;

[0214] A work order processing duration acquisition module 340, configured to acquire the work order processing duration;

[0215] A work order priority determination module 350, configured to determine the work order priority according to the emotion analysis result, the level information, and the work order processing duration;

[0216] A dynamic allocation module 360, configured to perform engineer allocation on each work order in sequence based on the work order priority to obtain a dynamic allocation result of the engineer for each work order processed.

[0217] This device can execute the work order dynamic allocation method based on user emotion analysis provided by each of the above embodiments, and has corresponding functional units and beneficial effects. Details are not described herein again.

[0218] Embodiment 4

[0219] Those skilled in the art should understand that the embodiments of this solution can provide methods, systems, or computer program products. Therefore, this solution can be implemented in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, this solution can be implemented in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0220] Therefore, this application also proposes a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the method described in any one of the embodiments of this application.

[0221] This solution is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this solution. It should be understood that each process and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowcharts and / or block diagrams can also be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, so that the instructions executed by the processors of the computer or other programmable data processing devices generate a device for implementing the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0222] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more processes and / or blocks Figure 1 in the process Figure 1 or processes and / or blocks

[0223] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more processes and / or blocks Figure 1 in the process Figure 1 or processes and / or blocks

[0224] Further Figure 4 is a schematic structural diagram of an electronic device provided in Embodiment 4 of the present application. As Figure 4 shown, the present application also provides an electronic device (or computing device), including a processor 11, a memory 12, and a computer program stored on the memory 12 and executable on the processor 11, where the processor 11 implements the method according to any embodiment of the present application when executing the computer program

[0225] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory. The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media. Computer-readable media includes both permanent and non-permanent, removable and non-removable media implemented by any method or technology for information storage. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0226] It should also be noted that the term "comprising", "including", or any other variation thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the statement "including a..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0227] The above description is only for the embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A method for dynamically allocating work orders based on user sentiment analysis, characterized in that, The method includes: Obtaining multi-modal information of the work orders submitted by users; wherein, the multi-modal information includes one or more of work order description text information, user voice information, and submission behavior information; Performing user emotion analysis based on the multi-modal information to obtain an emotion analysis result; Obtaining the level information of the user; Obtaining the work order processing duration; Determining the work order priority according to the emotion analysis result, the level information, and the work order processing duration; Performing engineer allocation on each work order in sequence based on the work order priority to obtain a dynamic allocation result of the engineer for each work order processed; 2. The method according to claim 1, characterized in that, Performing engineer allocation on each work order in sequence based on the work order priority to obtain a dynamic allocation result of the engineer for each work order processed, including: Determining the work orders to be allocated in sequence based on the work order priority, in the order from high to low priority; Obtaining the static skill information of the engineer and the dynamic matching information for the work orders to be allocated; Generating a list of engineer matching degrees for the work orders to be allocated according to the static skill information and the dynamic matching information; Determining the target engineer for the work orders to be allocated based on the sorting of the engineers in the list of engineer matching degrees, and traversing all work orders to obtain a dynamic allocation result of the engineer for each work order processed.

3. The method according to claim 1, characterized in that, Determining the work order priority according to the emotion analysis result, the level information, and the work order processing duration, including: Obtaining the emotion score corresponding to the emotion analysis result, the level score corresponding to the level information, and the duration score corresponding to the work order processing duration; Obtaining the emotion weight value, the level weight value, and the processing duration weight value; Determining the priority score of the work order according to the product of the emotion score and the emotion weight value, the product of the level score and the level weight value, and the product of the duration score and the processing duration weight value; Determining the work order priority according to the priority scores of all work orders.

4. The method according to claim 3, wherein The emotion score is determined based on the interval in which the emotion analysis result falls; The level score is determined based on the interval in which the level information falls; The duration score is determined based on the interval in which the work order processing duration falls.

5. The method according to claim 1, characterized in that Performing user emotion analysis based on the multi-modal information to obtain an emotion analysis result, including: Performing user emotion analysis according to the number of keywords and punctuation mark features in the work order description text information, the speech rate feature, pitch feature, and pause frequency feature of the user voice information, and the submission frequency feature and interaction path feature of the submission behavior in the multi-modal information to obtain an emotion analysis result.

6. The method according to claim 1, wherein The method further includes: When a newly generated work order is recognized and the work order priority of the newly generated work order exceeds the set threshold, re-determining the work orders to be allocated and re-performing the dynamic allocation of engineers.

7. The method according to claim 1, wherein After obtaining the dynamic allocation result of the engineer for each work order processed, the method further includes: Obtaining the actual processing duration, engineer feedback, and user feedback of each work order, and adjusting the consideration factors for engineer allocation.

8. A work order dynamic allocation device based on user emotion analysis, characterized in that, The device includes: A multimodal information acquisition module, configured to acquire multimodal information of a work order submitted by a user; wherein, the multimodal information includes one or more of work order description text information, user voice information, and submission behavior information; An emotion analysis result determination module, configured to perform user emotion analysis based on the multimodal information to obtain an emotion analysis result; A level information acquisition module, configured to acquire the level information of the user; A work order processing duration acquisition module, configured to acquire the work order processing duration; A work order priority determination module, configured to determine the work order priority according to the emotion analysis result, the level information, and the work order processing duration; A dynamic allocation module, configured to perform engineer allocation for each work order in sequence based on the work order priority to obtain a dynamic allocation result of the engineer for each work order processing.

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

10. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method according to any one of claims 1-7.

Images