Intelligent customer service telephone repair method and system based on AI

Abstract

The invention discloses an intelligent customer service telephone repair method and system based on AI (artificial intelligence), and relates to the technical field of artificial intelligence, and the method comprises the following steps: step 1, information input, step 2, information processing, step 3, fault analysis, step 4, work order generation and distribution, and step 5, knowledge base updating. The repair type of the incoming call user is obtained, the repair emergency degree of the incoming call user is analyzed from multiple aspects, initiation conditions corresponding to the air conditioner repair type are obtained, the accuracy of fault analysis is improved, appropriate engineers are screened according to multiple elements to conduct door-to-door maintenance service, the accuracy of order receiving of the engineers is improved, and the efficiency is improved. Finally, the storage demand coefficient of the work order of the incoming call user is evaluated, the work order of the incoming call user is processed, a knowledge base is enriched, and redundancy of the knowledge base is avoided.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an AI-based intelligent customer service telephone repair reporting method and system. Background Art

[0002] With the rapid development of information technology and network technology, as well as the increasing requirements of users for service efficiency and quality, the traditional manual customer service model has been unable to meet the needs of modern society. In this context, intelligent customer service systems have emerged and gradually become an important way to solve customer problems, such as air conditioning telephone repair. Traditional customer service is less professional and inefficient, and cannot provide correct solutions based on the type of repair requested by the caller. When users encounter problems, they are usually required to go to a fixed maintenance point for equipment repair, which undoubtedly increases the inconvenience and time cost of users. Therefore, it is particularly important to develop an AI-based intelligent customer service telephone repair method and system.

[0003] The existing intelligent customer service telephone repair service still has the following defects, which are specifically reflected in:

[0004] 1. The existing intelligent customer service telephone repair reporting has limitations and poor professionalism. It rarely analyzes the urgency of the caller's air conditioning repair report from multiple levels, which affects the subsequent determination of the caller's appropriate waiting time based on the urgency of the caller's air conditioning repair report, thereby reducing the compatibility between the caller and the dispatch engineer, and reducing the caller's repair efficiency and maintenance effect.

[0005] 2. The existing intelligent customer service telephone repair reporting system often randomly arranges engineers to perform repairs when processing the repair reports of callers. Engineers accept orders blindly and are unable to select the most suitable engineers for repairs based on their historical service work orders, which reduces the intelligent repair reporting experience of callers and thus reduces the value of the intelligent customer service telephone repair reporting platform. Summary of the invention

[0006] The purpose of the present invention is to provide an AI-based intelligent customer service telephone repair reporting method and system to solve the problems existing in the background technology.

[0007] To solve the above technical problems, the present invention adopts the following technical solutions: an AI-based intelligent customer service telephone repair method, comprising: step one, information entry: AI intelligent customer service answers the caller's call, records the caller's voice, and invites the caller to record a video.

[0008] Step 2: Information processing: Based on the caller's voice and the video uploaded by the caller, the caller's repair type and its corresponding repair urgency are evaluated.

[0009] Step 3, fault analysis: according to the repair type of the caller, obtain the triggering conditions corresponding to the repair type of the caller's air conditioner.

[0010] Step 4: Work order generation and dispatch: Based on the triggering conditions corresponding to the caller's repair report type, obtain the historical maintenance records of each engineer, each dispatched work order and its corresponding location information from the AI ​​customer service center, screen the appropriate engineer for the caller, generate the caller's work order, and dispatch it to the caller and the caller's appropriate engineer respectively.

[0011] Step 5, knowledge base update: When the caller's work order is completed, the caller's appropriate engineer will update the maintenance record section of the work order, regenerate the caller's service record work order, evaluate the caller's work order storage requirement coefficient based on the caller's service record work order, and process the caller's work order.

[0012] As a preferred solution, the specific analysis method for evaluating the repair type of the incoming caller is as follows:

[0013] Convert the caller's voice into text, obtain the keywords of the caller's voice text, construct the caller's voice text keyword set R', and obtain the keyword set R corresponding to each repair type from the data i , evaluate the similarity between the caller's voice text keyword set and the keyword set of each repair type, the calculation formula is, Among them J i R represents the similarity between the caller's voice text keyword set and the keyword set of the i-th repair type, i A keyword set representing the i-th repair report type is used, and the repair report type corresponding to the maximum similarity value is used as the repair report type of the incoming caller. i represents the repair report type number, i=1,2,...,j, and j is a positive integer greater than 2.

[0014] As a preferred solution, the specific analysis method for evaluating the caller's repair urgency is as follows:

[0015] Based on the video uploaded by the caller, the loudness value of the air conditioner's external fan and the area of ​​water splashing caused by the air conditioner leakage in the video are obtained. With the user's authorization, the volume of the insulation layer of the air conditioner's external copper pipe, the temperature of the air conditioner's outlet, the air conditioner's preset temperature and the air conditioner's operating voltage are obtained.

[0016] The standard loudness value of the external fan of the air conditioner, the allowable dripping area of ​​the air conditioner and the initial volume of the insulation layer of the external copper pipe connecting the air conditioner of the caller are obtained from the database.

[0017] Get the speaking speed V′ of the caller, and calculate the urgency of the speaking speed of the caller according to the speaking speed of the caller. The calculation formula is: Where α represents the urgency of the caller's speaking speed, V′ represents the caller's speaking speed, V represents the standard speaking speed, and e represents a natural constant.

[0018] Based on the urgency of the caller's speech speed, if the air conditioner repair type is dripping, the repair urgency of the caller's air conditioner dripping is calculated using the following formula: Where δ represents the urgency of the air conditioner dripping splashing area, S′ represents the air conditioner dripping splashing area obtained from the video, and S represents the allowable dripping area of ​​the air conditioner.

[0019] If the air conditioner repair type is abnormal noise from the air conditioner external fan, the repair urgency of the caller's air conditioner external fan abnormal noise is calculated using the following formula: Where ε represents the urgency of the abnormal sound of the air conditioner external fan, dB′ represents the loudness value of the air conditioner external fan obtained from the video, and dB represents the standard loudness value of the air conditioner external fan.

[0020] If the air conditioner repair type is air conditioner refrigeration abnormality, the repair urgency of the caller's air conditioner refrigeration abnormality is calculated using the following formula: Where η represents the urgency of the abnormal cooling of the air conditioner, T′ represents the temperature of the air outlet of the air conditioner, T represents the preset temperature of the air conditioner, M′ represents the operating voltage of the air conditioner, M represents the rated voltage of the air conditioner, N′ represents the volume of the insulation layer of the copper pipe connected to the outside of the air conditioner, and N represents the initial volume of the insulation layer of the copper pipe connected to the outside of the air conditioner of the caller.

[0021] The repair urgency β of the incoming caller is obtained, where β is δ, ε or η.

[0022] As a preferred solution, the specific analysis method for screening the appropriate engineers for the incoming caller is as follows:

[0023] Analyze the initial compatibility between the incoming caller and each engineer, and obtain the initial adaptation engineers for the incoming caller based on the initial compatibility. Then screen and obtain the intermediate adaptation engineers for the incoming caller, evaluate the suitability dispatch evaluation index of each intermediate adaptation engineer, and obtain the suitable engineer for the incoming caller.

[0024] As a preferred solution, the initial compatibility between the incoming caller and each engineer is analyzed by the following specific analysis method: the historical maintenance records of each engineer, each dispatched work order and its corresponding location information are obtained from the AI ​​customer service center, wherein the historical maintenance records include the repair type of each historically dispatched work order, the triggering conditions and maintenance duration corresponding to each repair type.

[0025] According to the repair report type and triggering conditions of each engineer in each historical work order dispatch, the historical work orders of each repair report type and the historical work orders of each triggering condition of each engineer are obtained, and the total number of historical dispatched work orders, the historical processing times of each repair report type and the historical processing times of each triggering condition of each engineer are obtained by statistics.

[0026] According to the repair type of the caller, record it as the target repair type, filter the historical processing times of each engineer on the target repair type, count the historical processing times of each engineer on the target repair type, and count the historical processing times of each engineer on all repair types.

[0027] According to the triggering conditions of the corresponding repair report type of the calling user, they are recorded as target triggering conditions, the historical processing times of each engineer on each target triggering condition are screened, the historical processing times of each engineer on all target triggering conditions are counted, and the historical processing times of each engineer on all triggering conditions are counted.

[0028] Analyze the initial compatibility between the caller and each engineer. The calculation formula is: where ψ p represents the initial fit between the caller and the pth engineer, a′ p represents the historical number of times the pth engineer has processed the target repair type, a p represents the historical number of times the pth engineer has processed all types of repairs, b p represents the number of times the pth engineer has handled all triggering conditions, b′ p It represents the number of times the p-th engineer has processed all target triggering conditions in history. p represents the number of each engineer. p=1,2,...,r, where r is a positive integer greater than 2.

[0029] As a preferred solution, the specific analysis method for the intermediate adaptation engineers of the incoming callers is as follows: based on the urgency of the caller's repair report, combined with the suitable waiting time corresponding to each repair urgency interval stored in the database, the suitable waiting time for the incoming caller is screened, and the dispatched work orders of the initial adaptation engineers of the incoming caller are obtained, and then the waiting time for each initial adaptation engineer of the incoming caller is extracted.

[0030] The appropriate waiting time of the incoming caller is compared with the waiting time of each initial adaptation engineer. If the appropriate waiting time of the incoming caller is longer than the waiting time of a certain initial adaptation engineer, the initial adaptation engineer is recorded as an intermediate adaptation engineer, and the intermediate adaptation engineers of the incoming caller are obtained.

[0031] As a preferred solution, the evaluation index of the suitability of dispatching orders of each intermediate adaptation engineer is evaluated by a specific analysis method: according to the intermediate adaptation engineers of the calling user, combined with the maintenance time of each historical dispatched work order of each intermediate adaptation engineer, the final location of the dispatched work order of each intermediate adaptation engineer, and the location of the calling user, the evaluation index of the suitability of dispatching orders of each intermediate adaptation engineer is evaluated, and the calculation formula is, where γ q represents the suitability evaluation index of the qth intermediate adaptation engineer, μ q represents the maintenance time of the qth intermediate adaptation engineer, θ q represents the distance between the final location of the dispatched work order of the qth intermediate adaptation engineer and the calling user, z represents the number of intermediate adaptation engineers, q represents the number of each intermediate adaptation engineer, q=1,2,...,z, where z is a positive integer greater than 2.

[0032] As a preferred solution, the storage requirement coefficient of the service work order of the incoming caller is evaluated, and the specific analysis method is: according to the service work order of the incoming caller, the repair type of the incoming caller and its corresponding triggering conditions and maintenance parameters are obtained, wherein the maintenance parameters include the tools used, the replacement parts, and the maintenance time.

[0033] Each historical service work order is extracted from the database, and the repair type and its corresponding triggering conditions and maintenance parameters are extracted therefrom. According to the keywords of the repair type and its corresponding triggering conditions of each historical service work order, a keyword set of the triggering conditions corresponding to the repair type of each historical service work order is constructed.

[0034] According to the repair type of the service work order of the incoming user, combined with the keyword set of the triggering conditions corresponding to the repair type of each historical service work order, if the keyword set of the triggering conditions corresponding to the repair type of the service work order of the incoming user cannot be extracted, the storage demand coefficient of the service work order of the incoming user is recorded as U. If the keyword set of the triggering conditions corresponding to the repair type of the service work order of the incoming user is extracted, the following analysis is performed: Calculate the repair similarity between the service work order of the incoming user and each historical service work order, and the calculation formula is, Among them A x represents the similarity between the service ticket of the incoming user and the repair report of the xth historical service ticket, B represents the keyword set of the triggering condition corresponding to the repair type of the service ticket of the incoming user, and C x A keyword set representing the triggering condition corresponding to the repair type of the x-th historical service work order, where x represents the number of each historical service work order, x=1,2,...,y, and y is a positive integer greater than 2.

[0035] If the repair similarity between the incoming caller's service order and each historical service order is less than the preset repair similarity threshold, the storage requirement coefficient of the incoming caller's service order is recorded as Y. Otherwise, the repair similarity between the incoming caller's service order and each associated historical service order is analyzed.

[0036] If the maintenance similarity between the service work order of the incoming user and a certain historical service work order is less than a preset maintenance similarity threshold, the storage demand coefficient of the service work order of the incoming user is recorded as T, otherwise, it is recorded as W.

[0037] The storage requirement coefficient G of the service ticket of the incoming caller is obtained, where the value of G is U, Y, T or W.

[0038] As a preferred scheme, the maintenance similarity between the service work order of the incoming caller and the related historical service work orders is evaluated. The specific analysis method is: the repair similarity between the service work order of the incoming caller and each historical service work order is compared with a preset repair similarity threshold. If the repair similarity between the service work order of the incoming caller and a historical service work order is greater than the preset repair similarity threshold, then the historical service work order is recorded as a related historical service work order, and the related historical service work orders corresponding to the incoming caller are obtained.

[0039] Based on each associated historical service work order corresponding to the calling user, each used tool, each replacement component, and maintenance duration in the maintenance parameters corresponding to each associated historical service work order of the calling user are obtained.

[0040] According to the tools and replacement parts in the maintenance parameters of the caller's service work order, the same tools and replacement parts between the caller's service work order and the associated historical service work orders are filtered, and the number of the same tools and the number of the same replacement parts are counted.

[0041] Evaluate the repair similarity between the caller's service ticket and each associated historical service ticket, Where T k represents the maintenance similarity between the caller's service ticket and the kth associated historical service ticket, H′ represents the number of tools used in the maintenance parameters of the caller's service ticket, and H k represents the number of identical tools used in the maintenance parameters of the kth associated historical service work order, V′ represents the number of replacement parts in the maintenance parameters of the caller’s service work order, and V k represents the number of identical replacement parts in the maintenance parameters of the kth associated historical service work order, SC′ represents the maintenance duration in the maintenance parameters of the service work order of the calling user, and SC k represents the maintenance duration of the kth related historical service work order, k represents the number of each related historical service work order, k=1,2,...,t, t is a positive integer greater than 2.

[0042] The second aspect of the present invention provides a system for an AI-based intelligent customer service telephone repair method, comprising: an information entry module, in which the AI ​​intelligent customer service answers the caller's call, records the caller's voice, and invites the caller to record a video.

[0043] The information processing module evaluates the caller's repair type and its corresponding repair urgency based on the caller's voice and the video uploaded by the caller.

[0044] The fault analysis module obtains various triggering conditions corresponding to the repair type of the calling user's air conditioner according to the repair type of the calling user.

[0045] The work order generation and dispatching module obtains the historical maintenance records of each engineer, each dispatched work order and its corresponding location information from the AI ​​customer service center according to the triggering conditions corresponding to the caller's repair report type, screens the appropriate engineer for the caller, generates the caller's work order, and dispatches it to the caller and the caller's appropriate engineer respectively.

[0046] Knowledge base update module: when the caller's work order is completed, the caller's appropriate engineer will update the maintenance record section of the work order, regenerate the caller's service record work order, evaluate the caller's work order storage requirement coefficient based on the caller's service record work order, and process the caller's work order.

[0047] Compared with the prior art, the beneficial effects of the present invention are: 1. The intelligent customer service telephone repair reporting in the present invention has improved accuracy, and analyzes the urgency of the caller's air-conditioning repair report from multiple aspects such as the caller's speech speed, air-conditioning dripping, abnormal sound of the air-conditioning external fan and air-conditioning refrigeration abnormality, and determines the appropriate waiting time for the caller according to the urgency of the caller's air-conditioning repair report, quickly determines the compatibility between the caller and the dispatch engineer, and enhances the caller's repair efficiency and maintenance effect.

[0048] 2. The intelligent customer service telephone repair reporting in the present invention can conduct a comprehensive analysis from multiple aspects based on the location information of the caller, screen out the most suitable engineer for on-site repair service, improve the accuracy of engineers taking orders, and improve the caller's experience of intelligent customer service telephone repair reporting, thereby raising the value of building an intelligent customer service telephone repair reporting platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 The present invention is a schematic flow chart of the steps for implementing the method.

[0051] Figure 2 It is a schematic diagram of the system module of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] Reference Figure 1 As shown, the present invention provides an AI-based intelligent customer service telephone repair method, including: step one, information entry: the AI ​​intelligent customer service answers the caller's call, records the caller's voice, and invites the caller to record a video.

[0054] Step 2: Information processing: Based on the caller's voice and the video uploaded by the caller, the caller's repair type and its corresponding repair urgency are evaluated.

[0055] In a specific embodiment, the specific analysis method for evaluating the repair type of the incoming caller is as follows:

[0056] Convert the caller's voice into text, obtain the keywords of the caller's voice text, construct the caller's voice text keyword set R', and obtain the keyword set R corresponding to each repair type from the data i , evaluate the similarity between the caller's voice text keyword set and the keyword set of each repair type, the calculation formula is, Among them J i R represents the similarity between the caller's voice text keyword set and the keyword set of the i-th repair type, i A keyword set representing the i-th repair report type is used, and the repair report type corresponding to the maximum similarity value is used as the repair report type of the incoming caller. i represents the repair report type number, i=1,2,...,j, and j is a positive integer greater than 2.

[0057] In a specific embodiment, the specific analysis method for evaluating the caller's repair urgency is as follows:

[0058] Based on the video uploaded by the caller, the loudness value of the air conditioner's external fan and the area of ​​water splashing caused by the air conditioner leakage in the video are obtained. With the user's authorization, the volume of the insulation layer of the air conditioner's external copper pipe, the temperature of the air conditioner's outlet, the air conditioner's preset temperature and the air conditioner's operating voltage are obtained.

[0059] The standard loudness value of the external fan of the air conditioner, the allowable dripping area of ​​the air conditioner and the initial volume of the insulation layer of the external copper pipe connecting the air conditioner of the caller are obtained from the database.

[0060] Get the speaking speed V′ of the caller, and calculate the urgency of the speaking speed of the caller according to the speaking speed of the caller. The calculation formula is: Where α represents the urgency of the caller's speaking speed, V′ represents the caller's speaking speed, V represents the standard speaking speed, and e represents a natural constant.

[0061] It should be noted that the standard speaking speed is about 240 characters per minute.

[0062] Based on the urgency of the caller's speech speed, if the air conditioner repair type is dripping, the repair urgency of the caller's air conditioner dripping is calculated using the following formula: Where δ represents the urgency of the air conditioner dripping splashing area, S′ represents the air conditioner dripping splashing area obtained from the video, and S represents the allowable dripping area of ​​the air conditioner.

[0063] It should be noted that the allowable dripping area of ​​the air conditioner is determined by the engineer and is set to 1cm here. 2 .

[0064] If the air conditioner repair type is abnormal noise from the air conditioner external fan, the repair urgency of the caller's air conditioner external fan abnormal noise is calculated using the following formula: Where ε represents the urgency of the abnormal sound of the air conditioner external fan, dB′ represents the loudness value of the air conditioner external fan obtained from the video, and dB represents the standard loudness value of the air conditioner external fan.

[0065] It should be noted that the standard loudness value of the air conditioner's external fan depends on the environment, and the value here is 58 decibels.

[0066] If the air conditioner repair type is air conditioner refrigeration abnormality, the repair urgency of the caller's air conditioner refrigeration abnormality is calculated using the following formula: Where η represents the urgency of the abnormal cooling of the air conditioner, T′ represents the temperature of the air outlet of the air conditioner, T represents the preset temperature of the air conditioner, M′ represents the operating voltage of the air conditioner, M represents the rated voltage of the air conditioner, N′ represents the volume of the insulation layer of the copper pipe connected to the outside of the air conditioner, and N represents the initial volume of the insulation layer of the copper pipe connected to the outside of the air conditioner of the caller.

[0067] It should be noted that the operating voltage of the air conditioner is 220V, and the initial volume of the insulation layer of the external copper pipe connected to the air conditioner indicates the volume of the insulation layer of the external copper pipe connected to the air conditioner when the air conditioner is first installed.

[0068] The repair urgency β of the incoming caller is obtained, where β is δ, ε or η.

[0069] The intelligent customer service telephone repair reporting in the present invention has improved accuracy, analyzes the urgency of the caller's air-conditioning repair report from multiple aspects such as the caller's speech speed, air-conditioner dripping, abnormal sound of the air-conditioner external fan and air-conditioner refrigeration anomaly, determines the appropriate waiting time for the caller according to the urgency of the caller's air-conditioning repair report, quickly determines the compatibility between the caller and the dispatch engineer, and enhances the caller's repair reporting efficiency and maintenance effect.

[0070] Step 3, fault analysis: according to the repair type of the caller, obtain the triggering conditions corresponding to the repair type of the caller's air conditioner.

[0071] Obtain each service ticket from the AI ​​customer service center, and extract each fault type and triggering condition from it. According to the fault type of the caller's air conditioner and the video uploaded by the caller, filter the triggering conditions corresponding to the air conditioner fault type.

[0072] It should be noted that if the fault type is dripping air conditioner, the corresponding triggering conditions are equipment problems, drainage and pipeline problems, etc. If the fault type is abnormal noise from the external fan of the air conditioner, the corresponding triggering conditions are damage or looseness of the fan blades, fan motor failure and foreign objects entering the fan, etc. If the fault type is abnormal air conditioning refrigeration, the corresponding triggering conditions are refrigerant leakage or shortage, dirty and blocked filter and poor heat dissipation of the condenser, etc.

[0073] Step 4: Work order generation and dispatch: Based on the triggering conditions corresponding to the caller's repair report type, obtain the historical maintenance records of each engineer, each dispatched work order and its corresponding location information from the AI ​​customer service center, screen the appropriate engineer for the caller, generate the caller's work order, and dispatch it to the caller and the caller's appropriate engineer respectively.

[0074] In a specific embodiment, the specific analysis method for screening suitable engineers for incoming callers is as follows: analyzing the initial compatibility between the incoming caller and each engineer, obtaining each initial adaptation engineer for the incoming caller based on the initial compatibility, and then screening each intermediate adaptation engineer for the incoming caller, evaluating the suitability dispatch evaluation index of each intermediate adaptation engineer, and obtaining a suitable engineer for the incoming caller.

[0075] In a specific embodiment, the initial compatibility between the incoming caller and each engineer is analyzed by the following specific analysis method: the historical maintenance records of each engineer, each dispatched work order and its corresponding location information are obtained from the AI ​​customer service center, wherein the historical maintenance records include the repair type of each historically dispatched work order, the triggering conditions and maintenance duration corresponding to each repair type.

[0076] According to the repair report type and triggering conditions of each engineer in each historical work order dispatch, the historical work orders of each repair report type and the historical work orders of each triggering condition of each engineer are obtained, and the total number of historical dispatched work orders, the historical processing times of each repair report type and the historical processing times of each triggering condition of each engineer are obtained by statistics.

[0077] According to the repair type of the caller, record it as the target repair type, filter the historical processing times of each engineer on the target repair type, count the historical processing times of each engineer on the target repair type, and count the historical processing times of each engineer on all repair types.

[0078] According to the triggering conditions of the corresponding repair report type of the calling user, they are recorded as target triggering conditions, the historical processing times of each engineer on each target triggering condition are screened, the historical processing times of each engineer on all target triggering conditions are counted, and the historical processing times of each engineer on all triggering conditions are counted.

[0079] Analyze the initial compatibility between the caller and each engineer. The calculation formula is: where ψ p represents the initial fit between the caller and the pth engineer, a′ p represents the historical number of times the pth engineer has processed the target repair type, a p represents the historical number of times the pth engineer has processed all types of repairs, b p represents the historical number of times the pth engineer handles all triggering conditions, b′ p It represents the number of times the p-th engineer has processed all target triggering conditions in history. p represents the number of each engineer. p=1,2,...,r, where r is a positive integer greater than 2.

[0080] In a specific embodiment, the specific analysis method for the intermediate adaptation engineers of the incoming caller is as follows: based on the urgency of the caller's repair report, combined with the suitable waiting time corresponding to each repair urgency interval stored in the database, the suitable waiting time for the incoming caller is screened, and the dispatched work orders of the initial adaptation engineers of the incoming caller are obtained, and then the waiting time for each initial adaptation engineer of the incoming caller is extracted.

[0081] It should be noted that the pending processing time represents the total processing time of each work order that has been dispatched by the engineer.

[0082] The appropriate waiting time of the incoming caller is compared with the waiting time of each initial adaptation engineer. If the appropriate waiting time of the incoming caller is longer than the waiting time of a certain initial adaptation engineer, the initial adaptation engineer is recorded as an intermediate adaptation engineer, and the intermediate adaptation engineers of the incoming caller are obtained.

[0083] In a specific embodiment, the evaluation method for evaluating the appropriate dispatching evaluation index of each intermediate adaptation engineer is as follows: based on the intermediate adaptation engineers of the caller, combined with the maintenance time of each historical dispatched work order of each intermediate adaptation engineer, the final location of the dispatched work order of each intermediate adaptation engineer, and the location of the caller, the evaluation index of the appropriate dispatching evaluation index of each intermediate adaptation engineer is evaluated, and the calculation formula is: where γ q represents the suitability evaluation index of the qth intermediate adaptation engineer, μ q represents the maintenance time of the qth intermediate adaptation engineer, θ q represents the distance between the final location of the dispatched work order of the qth intermediate adaptation engineer and the calling user, z represents the number of intermediate adaptation engineers, q represents the number of each intermediate adaptation engineer, q=1,2,...,z, where z is a positive integer greater than 2.

[0084] It should be noted that the final location indicates the location of the last work order that the engineer has issued.

[0085] It should be noted that the specific analysis method for the suitable engineer for the incoming caller is as follows: the engineer corresponding to the maximum value of the suitable dispatch evaluation index is taken as the suitable engineer for the incoming caller.

[0086] The intelligent customer service telephone repair reporting in the present invention can conduct a comprehensive analysis from multiple aspects based on the location information of the caller, screen out the most suitable engineer for on-site repair service, improve the accuracy of engineers taking orders, and improve the caller's experience of intelligent customer service telephone repair reporting, thereby raising the value of building an intelligent customer service telephone repair reporting platform.

[0087] Step 5, knowledge base update: When the caller's work order is completed, the caller's appropriate engineer will update the maintenance record section of the work order, regenerate the caller's service record work order, evaluate the caller's work order storage requirement coefficient based on the caller's service record work order, and process the caller's work order.

[0088] In a specific embodiment, the storage requirement coefficient of the service work order of the evaluating caller is specifically analyzed by: according to the service work order of the calling user, the repair type of the calling user and its corresponding triggering conditions and maintenance parameters are obtained, wherein the maintenance parameters include various tools to be used, various replacement parts, and maintenance time.

[0089] Each historical service work order is extracted from the database, and the repair type and its corresponding triggering conditions and maintenance parameters are extracted therefrom. According to the keywords of the repair type and its corresponding triggering conditions of each historical service work order, a keyword set of the triggering conditions corresponding to the repair type of each historical service work order is constructed.

[0090] According to the repair type of the service ticket of the incoming user, combined with the keyword set of the triggering conditions corresponding to the repair type of each historical service ticket, if the keyword set of the triggering conditions corresponding to the repair type of the service ticket of the incoming user cannot be extracted, the storage demand coefficient of the service ticket of the incoming user is recorded as U. If the keyword set of the triggering conditions corresponding to the repair type of the service ticket of the incoming user is extracted, the following analysis is performed:

[0091] Calculate the similarity between the caller's service ticket and each historical service ticket. The calculation formula is: Among them A x represents the similarity between the service ticket of the incoming user and the repair report of the xth historical service ticket, B represents the keyword set of the triggering condition corresponding to the repair type of the service ticket of the incoming user, and C x A keyword set representing the triggering condition corresponding to the repair type of the x-th historical service work order, where x represents the number of each historical service work order, x=1,2,...,y, and y is a positive integer greater than 2.

[0092] If the repair similarity between the incoming caller's service order and each historical service order is less than the preset repair similarity threshold, the storage requirement coefficient of the incoming caller's service order is recorded as Y. Otherwise, the repair similarity between the incoming caller's service order and each associated historical service order is analyzed.

[0093] If the maintenance similarity between the service work order of the incoming user and a certain historical service work order is less than a preset maintenance similarity threshold, the storage demand coefficient of the service work order of the incoming user is recorded as T, otherwise, it is recorded as W.

[0094] The storage requirement coefficient G of the service ticket of the incoming caller is obtained, where the value of G is U, Y, T or W.

[0095] It should be noted that the specific method of processing the work order of the incoming user is as follows: if the storage demand coefficient G of the service work order of the incoming user takes the values ​​of U, Y and T, the service work order of the incoming user is stored; otherwise, the service work order of the incoming user is stored in the buffer database.

[0096] In a specific embodiment, the evaluation of the maintenance similarity between the service work order of the incoming caller and each associated historical service work order is performed by comparing the repair similarity between the service work order of the incoming caller and each historical service work order with a preset repair similarity threshold. If the repair similarity between the service work order of the incoming caller and a historical service work order is greater than the preset repair similarity threshold, then the historical service work order is recorded as an associated historical service work order, and the associated historical service work orders corresponding to the incoming caller are obtained.

[0097] Based on each associated historical service work order corresponding to the calling user, each used tool, each replacement component, and maintenance duration in the maintenance parameters corresponding to each associated historical service work order of the calling user are obtained.

[0098] According to the tools and replacement parts in the maintenance parameters of the caller's service work order, the same tools and replacement parts between the caller's service work order and the associated historical service work orders are filtered, and the number of the same tools and the number of the same replacement parts are counted.

[0099] Evaluate the repair similarity between the caller's service ticket and each associated historical service ticket, Where T k represents the maintenance similarity between the caller's service ticket and the kth associated historical service ticket, H′ represents the number of tools used in the maintenance parameters of the caller's service ticket, and H k represents the number of identical tools used in the maintenance parameters of the kth associated historical service work order, V′ represents the number of replacement parts in the maintenance parameters of the caller’s service work order, and V k represents the number of identical replacement parts in the maintenance parameters of the kth associated historical service work order, SC′ represents the maintenance duration in the maintenance parameters of the service work order of the calling user, and SC k represents the maintenance duration of the kth related historical service work order, k represents the number of each related historical service work order, k=1,2,...,t, t is a positive integer greater than 2.

[0100] Reference Figure 2 As shown, the second aspect of the present invention provides a system for an AI-based intelligent customer service telephone repair method, including: an information entry module, in which AI intelligent customer service answers the caller's call, records the caller's voice, and invites the caller to record a video.

[0101] The information processing module evaluates the caller's repair type and its corresponding repair urgency based on the caller's voice and the video uploaded by the caller.

[0102] The fault analysis module obtains various triggering conditions corresponding to the repair type of the calling user's air conditioner according to the repair type of the calling user.

[0103] The work order generation and dispatching module obtains the historical maintenance records of each engineer, each dispatched work order and its corresponding location information from the AI ​​customer service center according to the triggering conditions corresponding to the caller's repair report type, screens the appropriate engineer for the caller, generates the caller's work order, and dispatches it to the caller and the caller's appropriate engineer respectively.

[0104] Knowledge base update module: when the caller's work order is completed, the caller's appropriate engineer will update the maintenance record section of the work order, regenerate the caller's service record work order, evaluate the caller's work order storage requirement coefficient based on the caller's service record work order, and process the caller's work order.

[0105] The information processing module is connected to the information processing module, the information processing module is connected to the fault analysis module, the fault analysis module is connected to the work order generation and distribution module, the work order generation and distribution module is connected to the knowledge base update module, and the information processing module, the fault analysis module, the work order generation and distribution module and the knowledge base update module are also connected to the database.

[0106] It should be noted that the present invention also includes a database for storing the standard loudness value of the air conditioner's external fan, the allowable dripping area of ​​the air conditioner, the initial volume of the insulation layer of the caller's air conditioner's external connecting copper pipe, and the appropriate waiting time corresponding to each repair urgency interval and each historical service work order.

[0107] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. An AI-based intelligent customer service telephone repair method, characterized in that: include: Step 1: Information entry: AI intelligent customer service answers the caller's call, records the caller's voice, and invites the caller to record a video; Step 2: Information processing: Based on the voice of the caller and the video uploaded by the caller, the caller's repair type and the corresponding repair urgency are evaluated; Step 3, fault analysis: according to the repair type of the caller, obtain the triggering conditions corresponding to the repair type of the caller's air conditioner; Step 4: Work order generation and dispatch: According to the triggering conditions corresponding to the caller's repair type, obtain the historical maintenance records of each engineer, each dispatched work order and its corresponding location information from the AI ​​customer service center, select the appropriate engineer for the caller, generate the caller's work order, and dispatch it to the caller and the caller's appropriate engineer respectively; Step 5, knowledge base update: When the caller's work order is completed, the caller's appropriate engineer will update the maintenance record section of the work order, regenerate the caller's service record work order, evaluate the caller's work order storage requirement coefficient based on the caller's service record work order, and process the caller's work order.

2. According to claim 1, an AI-based intelligent customer service telephone repair method is characterized in that: The specific analysis method for evaluating the repair type of the incoming caller is as follows: Convert the caller's voice into text, obtain the keywords of the caller's voice text, construct the caller's voice text keyword set R', and obtain the keyword set R corresponding to each repair type from the data i , evaluate the similarity between the caller's voice text keyword set and the keyword set of each repair type, the calculation formula is, Among them J i R represents the similarity between the caller's voice text keyword set and the keyword set of the i-th repair type, i A keyword set representing the i-th repair report type is used, and the repair report type corresponding to the maximum similarity value is used as the repair report type of the incoming caller. i represents the repair report type number, i=1,2,...,j, and j is a positive integer greater than 2.

3. According to claim 2, an AI-based intelligent customer service telephone repair method is characterized in that: The specific analysis method for evaluating the caller's repair urgency is as follows: Based on the video uploaded by the caller, the loudness value of the air conditioner's external fan and the area of ​​the air conditioner's water splashing in the video are obtained. With the user's authorization, the volume of the insulation layer of the air conditioner's external copper pipe, the temperature of the air conditioner's air outlet, the air conditioner's preset temperature, and the air conditioner's operating voltage are obtained. Obtain the standard loudness value of the external fan of the air conditioner, the allowable dripping area of ​​the air conditioner, and the initial volume of the insulation layer of the copper pipe connecting the external air conditioner of the caller from the database; Get the speaking speed V′ of the caller, and calculate the urgency of the speaking speed of the caller according to the speaking speed of the caller. The calculation formula is: Where α represents the urgency of the caller's speaking speed, V′ represents the caller's speaking speed, V represents the standard speaking speed, and e represents a natural constant; Based on the urgency of the caller's speech speed, if the air conditioner repair type is dripping, the repair urgency of the caller's air conditioner dripping is calculated using the following formula: Where δ represents the urgency of the air conditioner dripping splashing area, S′ represents the air conditioner dripping splashing area obtained from the video, and S represents the air conditioner allowable dripping area; If the air conditioner repair type is abnormal noise from the air conditioner external fan, the repair urgency of the caller's air conditioner external fan abnormal noise is calculated using the following formula: Where ε represents the urgency of the abnormal sound of the air conditioner external fan, dB′ represents the loudness value of the air conditioner external fan obtained from the video, and dB represents the standard loudness value of the air conditioner external fan; If the air conditioner repair type is air conditioner refrigeration abnormality, the repair urgency of the caller's air conditioner refrigeration abnormality is calculated using the following formula: Where η represents the urgency of the abnormal cooling of the air conditioner, T′ represents the temperature of the air outlet of the air conditioner, T represents the preset temperature of the air conditioner, M′ represents the operating voltage of the air conditioner, M represents the rated voltage of the air conditioner, N′ represents the volume of the insulation layer of the copper pipe connected to the outside of the air conditioner, and N represents the initial volume of the insulation layer of the copper pipe connected to the outside of the air conditioner of the caller; The repair urgency β of the incoming caller is obtained, where β is δ, ε or η.

4. According to claim 1, the AI-based intelligent customer service telephone repair method is characterized in that: The specific analysis method for screening the suitable engineers for the incoming callers is as follows: Analyze the initial compatibility between the incoming caller and each engineer, and obtain the initial adaptation engineers for the incoming caller based on the initial compatibility. Then screen and obtain the intermediate adaptation engineers for the incoming caller, evaluate the suitability dispatch evaluation index of each intermediate adaptation engineer, and obtain the suitable engineer for the incoming caller.

5. The AI-based intelligent customer service telephone repair method according to claim 4, characterized in that: The specific analysis method for analyzing the initial compatibility between the calling user and each engineer is as follows: Obtain the historical maintenance records of each engineer, each dispatched work order and its corresponding location information from the AI ​​customer service center. The historical maintenance records include the repair type of each historically dispatched work order, the triggering conditions and maintenance duration corresponding to each repair type; According to the repair type and triggering conditions of each engineer in each historical dispatch work order, each historical dispatch work order of each repair type and each historical dispatch work order of each triggering condition of each engineer are obtained, and the total number of historical dispatch work orders, the historical processing times of each repair type and each triggering condition of each engineer are obtained by statistics; According to the repair type of the caller, record it as the target repair type, filter the historical processing times of each engineer in the target repair type, count the historical processing times of each engineer for the target repair type, and count the historical processing times of each engineer for all repair types; According to the triggering conditions of the caller's corresponding repair type, record them as target triggering conditions, screen the historical processing times of each engineer on each target triggering condition, count the historical processing times of each engineer on all target triggering conditions, and count the historical processing times of each engineer on all triggering conditions; Analyze the initial compatibility between the caller and each engineer. The calculation formula is: where ψ p represents the initial fit between the caller and the pth engineer, a′ p represents the historical number of times the pth engineer has processed the target repair type, a p represents the historical number of times the pth engineer has processed all types of repairs, b p represents the historical number of times the pth engineer handles all triggering conditions, b′ p It represents the number of times the p-th engineer has processed all target triggering conditions in history. p represents the number of each engineer. p=1,2,...,r, where r is a positive integer greater than 2.

6. The AI-based intelligent customer service telephone repair method according to claim 4, characterized in that: The specific analysis method for the intermediate adaptation engineers of the caller is as follows: According to the caller's repair urgency, combined with the appropriate waiting time corresponding to each repair urgency interval stored in the database, the caller's appropriate waiting time is screened, and the dispatched work orders of each initial adaptation engineer of the caller are obtained, and then the waiting time of each initial adaptation engineer of the caller is extracted; The appropriate waiting time of the incoming caller is compared with the waiting time of each initial adaptation engineer. If the appropriate waiting time of the incoming caller is longer than the waiting time of a certain initial adaptation engineer, the initial adaptation engineer is recorded as an intermediate adaptation engineer, and the intermediate adaptation engineers of the incoming caller are obtained.

7. The AI-based intelligent customer service telephone repair method according to claim 4, characterized in that: The specific analysis method for evaluating the suitability of dispatching orders for each intermediate adaptation engineer is as follows: According to the intermediate adaptation engineers of the caller, combined with the maintenance time of each historical dispatched work order of each intermediate adaptation engineer, the final location of the dispatched work order of each intermediate adaptation engineer, and the location of the caller, the appropriate dispatch evaluation index of each intermediate adaptation engineer is evaluated. The calculation formula is: where γ q represents the suitability evaluation index of the qth intermediate adaptation engineer, μ q represents the maintenance time of the qth intermediate adaptation engineer, θ q represents the distance between the final location of the dispatched work order of the qth intermediate adaptation engineer and the calling user, z represents the number of intermediate adaptation engineers, q represents the number of each intermediate adaptation engineer, q=1,2,...,z, where z is a positive integer greater than 2.

8. The AI-based intelligent customer service telephone repair method according to claim 1, characterized in that: The storage requirement coefficient of the service ticket of the evaluating calling user is specifically analyzed by: According to the service work order of the caller, the caller's repair type and its corresponding triggering conditions and repair parameters are obtained, wherein the repair parameters include the tools used, the replacement parts, and the repair time; Extract each historical service work order from the database, and extract the repair type and its corresponding triggering conditions and maintenance parameters therefrom, and construct a keyword set of the repair type corresponding to the triggering conditions of each historical service work order according to the keywords of the repair type and its corresponding triggering conditions of each historical service work order; According to the repair type of the service ticket of the incoming user, combined with the keyword set of the triggering conditions corresponding to the repair type of each historical service ticket, if the keyword set of the triggering conditions corresponding to the repair type of the service ticket of the incoming user cannot be extracted, the storage demand coefficient of the service ticket of the incoming user is recorded as U. If the keyword set of the triggering conditions corresponding to the repair type of the service ticket of the incoming user is extracted, the following analysis is performed: Calculate the similarity between the caller's service ticket and each historical service ticket. The calculation formula is: Among them A x represents the similarity between the service ticket of the incoming user and the repair report of the xth historical service ticket, B represents the keyword set of the triggering condition corresponding to the repair type of the service ticket of the incoming user, and C x A keyword set representing the triggering condition corresponding to the repair type of the x-th historical service work order, where x represents the number of each historical service work order, x=1,2,...,y, where y is a positive integer greater than 2; If the repair similarity between the incoming caller's service order and each historical service order is less than the preset repair similarity threshold, the storage demand coefficient of the incoming caller's service order is recorded as Y. Otherwise, the repair similarity between the incoming caller's service order and each related historical service order is analyzed; If the maintenance similarity between the incoming user's service order and a certain historical service order is less than the preset maintenance similarity threshold, the storage demand coefficient of the incoming user's service order is recorded as T, otherwise, it is recorded as W; The storage requirement coefficient G of the service ticket of the incoming caller is obtained, where the value of G is U, Y, T or W.

9. The AI-based intelligent customer service telephone repair method according to claim 8, characterized in that: The specific analysis method for evaluating the maintenance similarity between the service ticket of the incoming caller and each associated historical service ticket is as follows: Compare the repair similarity between the incoming caller's service order and each historical service order with a preset repair similarity threshold. If the repair similarity between the incoming caller's service order and a certain historical service order is greater than the preset repair similarity threshold, the historical service order is recorded as an associated historical service order, and the associated historical service orders corresponding to the incoming caller are obtained; Based on each associated historical service work order corresponding to the calling user, obtaining each tool used, each replacement part, and maintenance duration in the maintenance parameters corresponding to each associated historical service work order of the calling user; According to the tools and replacement parts in the maintenance parameters of the service work order of the incoming user, the same tools and replacement parts between the service work order of the incoming user and the related historical service work orders are screened, and the number of the same tools and replacement parts is counted; Evaluate the repair similarity between the incoming user's service ticket and each associated historical service ticket, Where T k represents the maintenance similarity between the caller's service ticket and the kth associated historical service ticket, H′ represents the number of tools used in the maintenance parameters of the caller's service ticket, and H k represents the number of identical tools used in the maintenance parameters of the kth associated historical service work order, V′ represents the number of replacement parts in the maintenance parameters of the caller’s service work order, and V k represents the number of identical replacement parts in the maintenance parameters of the kth associated historical service work order, SC′ represents the maintenance duration in the maintenance parameters of the service work order of the calling user, and SC k represents the maintenance duration of the kth related historical service work order, k represents the number of each related historical service work order, k=1,2,...,t, t is a positive integer greater than 2.

10. A system for executing the AI-based intelligent customer service telephone repair method according to any one of claims 1 to 9, characterized in that: include: Information entry module, where AI intelligent customer service answers the caller's phone, records the caller's voice, and invites the caller to record a video; The information processing module evaluates the caller's repair type and its corresponding repair urgency based on the caller's voice and the video uploaded by the caller; The fault analysis module obtains the triggering conditions corresponding to the repair type of the caller's air conditioner according to the repair type of the caller; The work order generation and dispatching module obtains the historical maintenance records of each engineer, each dispatched work order and its corresponding location information from the AI ​​customer service center according to the triggering conditions corresponding to the caller's repair type, selects the appropriate engineer for the caller, generates the caller's work order, and dispatches it to the caller and the caller's appropriate engineer respectively; Knowledge base update module: when the caller's work order is completed, the caller's appropriate engineer will update the maintenance record section of the work order, regenerate the caller's service record work order, evaluate the caller's work order storage requirement coefficient based on the caller's service record work order, and process the caller's work order.

Images