CN116955574B - Intelligent customer service robot based on artificial intelligence and application method thereof - Google Patents
Intelligent customer service robot based on artificial intelligence and application method thereof Download PDFInfo
- Publication number
- CN116955574B CN116955574B CN202311208719.XA CN202311208719A CN116955574B CN 116955574 B CN116955574 B CN 116955574B CN 202311208719 A CN202311208719 A CN 202311208719A CN 116955574 B CN116955574 B CN 116955574B
- Authority
- CN
- China
- Prior art keywords
- customer service
- module
- preset
- query
- evaluation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 12
- 238000011156 evaluation Methods 0.000 claims abstract description 94
- 238000004458 analytical method Methods 0.000 claims abstract description 92
- 238000012545 processing Methods 0.000 claims abstract description 21
- 208000021663 Female sexual arousal disease Diseases 0.000 claims abstract description 19
- 208000006262 Psychological Sexual Dysfunctions Diseases 0.000 claims abstract description 19
- 230000006872 improvement Effects 0.000 claims abstract description 17
- 238000011835 investigation Methods 0.000 claims abstract description 13
- 238000001514 detection method Methods 0.000 claims description 32
- 230000003993 interaction Effects 0.000 claims description 22
- 230000004044 response Effects 0.000 claims description 13
- 238000012544 monitoring process Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 9
- 230000002452 interceptive effect Effects 0.000 claims description 6
- 230000001364 causal effect Effects 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 4
- 238000004064 recycling Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 12
- 230000009286 beneficial effect Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012797 qualification Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- Human Computer Interaction (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical field of intelligent robots, in particular to an intelligent customer service robot based on artificial intelligence and an application method thereof, wherein the intelligent customer service robot comprises a processor, an inquiry receiving and processing module, a knowledge base module, an intelligent reasoning and replying module, a problem distribution capturing and supervising module and a running performance evaluation and supervising module; according to the intelligent customer service robot, more accurate and humanized service can be provided for customers, the problem distribution of the customers is judged through analysis, the heat type and the frigidity type are determined, the frigidity type is analyzed to judge whether the customer service robot is of a failure answer type, so that a manager can update and perfect a knowledge base of the corresponding type of problems in a targeted manner, the operation performance of the intelligent customer service robot is evaluated and analyzed, and whether a investigation improvement signal is generated is judged through analysis when a qualified operation signal is generated, so that the use effect of the intelligent customer service robot is ensured, and the customer satisfaction is improved.
Description
Technical Field
The invention relates to the technical field of intelligent robots, in particular to an intelligent customer service robot based on artificial intelligence and an application method thereof.
Background
Intelligent customer service robots are novel robots that provide customer service through artificial intelligence technology, which can understand customer questions using natural language processing technology and provide automated answers through preset algorithms and rules, and can interact with customers in various ways, such as through sms, mail, social media, voice call, etc., and remain an important tool for enterprises to provide good customer service;
the traditional intelligent customer service robot is difficult to quickly, accurately and humanizedly provide service for clients, has low operation efficiency and poor service quality, cannot automatically analyze and determine problem distribution conditions, is not beneficial to timely updating and perfecting the content of corresponding types of problems in a knowledge base by management staff, cannot reasonably evaluate the operation of the intelligent customer service robot, is not beneficial to timely carrying out retrospective investigation and making corresponding improvement by the management staff, and is difficult to ensure the operation quality of the intelligent customer service robot;
in view of the above technical drawbacks, a solution is now proposed.
Disclosure of Invention
The invention aims to provide an intelligent customer service robot based on artificial intelligence and an application method thereof, which solve the problems that the prior art is difficult to quickly, accurately and humanizedly provide service for clients, the operation efficiency is low, the service quality is poor, the problem distribution condition cannot be automatically analyzed and determined, the operation of the intelligent customer service robot cannot be reasonably evaluated, and the operation quality of the intelligent customer service robot is difficult to ensure.
In order to achieve the above purpose, the present invention provides the following technical solutions:
an intelligent customer service robot based on artificial intelligence comprises a processor, an inquiry receiving and processing module, a knowledge base module, an intelligent reasoning and replying module, a problem distribution capturing and supervising module and a running performance evaluation and supervising module; the query receiving processing module is used for receiving natural language query of the client, understanding query content of the client, carrying out semantic analysis to identify the problem type of the client, and sending the query content and the identification analysis information to the intelligent reasoning reply module through the processor; the knowledge base module stores common questions and answers of clients, the intelligent reasoning reply module matches the questions of the clients with the existing questions in the knowledge base module one by one, and when the questions of the clients are matched with the existing questions in the knowledge base module, the corresponding answers are called from the knowledge base module and the clients are replied;
the problem distribution capturing and supervising module is used for collecting all problem types of customer service, judging the problem distribution of the customer through analysis, determining the heat type and the frigidity type, analyzing the frigidity type to judge whether the problem type is a solution failure type, generating a perfect updating signal of the solution failure type and the heat type, transmitting the perfect updating signal to a background monitoring pipe end through the processor, and updating and perfecting a knowledge base of the solution failure type and the heat type by a manager of the background monitoring pipe end; the operation performance evaluation module evaluates and analyzes the operation performance of the intelligent customer service robot so as to mark the corresponding detection time period as a high-quality operation time period or a low-quality operation time period, generates an operation qualified signal or an operation unqualified signal through analysis, and sends the operation qualified signal or the operation unqualified signal to the processor; the processor sends the disqualification signal to the background monitoring end.
Further, when no problem matched with the problem of the client exists in the knowledge base module, the intelligent reasoning reply module inquires related facts and information, deduces answers by applying a reasoning algorithm according to the inquired related facts and information, generates answers suitable for the client according to the reasoning result and replies to the client; the reasoning algorithm comprises logic reasoning, probability reasoning and causal reasoning.
Furthermore, the intelligent reasoning reply module is in communication connection with the man-machine interaction cooperation module, and realizes man-machine cooperation with human customer service personnel through the man-machine interaction cooperation module when the inquiry content of a client cannot be identified or the answer is uncertain, so that more accurate and more humanized service is provided, and specific operations comprise human customer service personnel intervention, man-machine common answer questions and human customer service personnel supervision;
the human customer service personnel intervenes in the method and the system for automatically transferring the problems to the human customer service personnel when the intelligent customer service robot cannot identify or determine the inquiry content of the customer; the man-machine jointly answer questions are used for jointly answering the questions of the clients by the human customer service personnel and the intelligent customer service robot when the questions of the clients are complex and require more information and knowledge to answer, the human customer service personnel provides more information and knowledge, and the intelligent customer service robot provides standardized answers to improve the accuracy and humanization degree of the service; the human customer service personnel supervision is used for supervising and evaluating the answers of the intelligent customer service robot by the human customer service personnel, and when the answers of the robot are found to be wrong or inaccurate, the human customer service personnel can intervene in time and correct the answers.
Further, the specific operation process of the problem distribution capturing and supervising module comprises the following steps:
collecting all problem types for customer service, marking the corresponding problem types as analysis objects i, i= {1,2, …, n }, wherein n represents the number of the problem types and n is a natural number greater than 1; acquiring the query frequency magnitude and the query number magnitude of the analysis object i, and carrying out numerical calculation on the query frequency magnitude and the query number magnitude to obtain a query coefficient; comparing the query coefficient with a preset query coefficient threshold value in a numerical mode, marking the corresponding analysis object i as a hot type if the query coefficient exceeds the preset query coefficient threshold value, and marking the corresponding analysis object i as a cold type if the query coefficient is not in the preset query coefficient threshold value;
if the analysis object i is of a frigidity type, collecting client feedback information of the analysis object i, wherein the client feedback information comprises solution accuracy of each query, carrying out numerical comparison on the solution accuracy and a preset solution accuracy threshold, and if the solution accuracy does not exceed the preset solution accuracy threshold, marking a solution result of the corresponding query as a non-qualified solution result; calculating the ratio of the generation times of the unqualified answer results to the query times to obtain unqualified answer values, and calculating the numerical value of the generation times of the unqualified answer results and the unqualified answer values to obtain answer evaluation values; and comparing the answer evaluation value with a preset answer evaluation threshold value, and if the answer evaluation value exceeds the preset answer evaluation threshold value, marking the corresponding frigidity type as an answer disqualification type.
Further, the specific operation process of the operation performance evaluation and supervision module comprises the following steps:
the method comprises the steps that response speeds and processing speeds of an intelligent customer service robot for each inquiry in a detection period are collected, all the response speeds in the detection period are summed, an average value is obtained to obtain a response average speed, all the processing speeds in the detection period are summed, and the average value is obtained to obtain a processing average speed; carrying out numerical calculation on the response average speed and the processing average speed to obtain an initial analysis value of the operation and assessment;
collecting online inquiring people butted by the intelligent customer service robot in a detection period, presetting a plurality of online inquiring people ranges, and distributing a group of preset initial analysis threshold values for the online inquiring people ranges; comparing the online inquired number with all the online inquired number ranges one by one to determine a preset initial analysis threshold value of the fortune assessment corresponding to the detection period; comparing the initial analysis value with a corresponding preset initial analysis threshold value, and marking the corresponding detection period as a low-quality operation period if the initial analysis value does not exceed the preset initial analysis threshold value; otherwise, the corresponding detection period is marked as a high quality operation period.
Further, after marking the corresponding detection time period as a low-quality operation time period or a high-quality operation time period, setting an operation evaluation period with a time length of K1, collecting the number of the high-quality operation time periods and the number of the low-quality operation time periods in the operation evaluation period, and calculating the ratio of the number of the low-quality operation time periods to the number of the high-quality operation time periods to obtain a low-quality operation parameter; performing numerical comparison on the low-quality operation parameter and a preset low-quality operation parameter threshold, generating an operation evaluation qualified signal if the low-quality operation parameter does not exceed the preset low-quality operation parameter threshold, and generating an operation evaluation unqualified signal if the low-quality operation parameter exceeds the preset low-quality operation parameter threshold; and sends the qualification signal or the disqualification signal to the processor.
Further, the processor is in communication connection with the questionnaire issuing and recovering module and the satisfaction decision feedback module, the processor sends the questionnaire issuing and recovering module when receiving the qualified operation and assessment signal, and the questionnaire sending and recovering module generates a questionnaire table of the intelligent customer service robot, sends the questionnaire table to a served customer and receives the questionnaire table fed back by the customer; and acquiring the starting answering time and the ending answering time of the corresponding clients, calculating the time difference between the ending answering time and the starting answering time to obtain an answering duration, comparing the answering duration with a preset answering duration, if the answering duration does not exceed the preset answering duration, removing the questionnaire list of the corresponding clients, marking the rest questionnaire list as a referenceable list, and transmitting all referenceable lists to a satisfaction decision feedback module through a processor.
Further, the specific operation process of the satisfaction decision feedback module comprises the following steps:
obtaining the evaluation results of each item in the corresponding referenceable table and assigning points to obtain interactive evaluation points of corresponding clients, summing all the interactive evaluation points and taking an average value to obtain an evaluation point representation value; comparing the evaluation score representation value with a preset evaluation score representation threshold value in a numerical value manner, and generating a investigation improvement signal if the evaluation score representation value does not exceed the preset evaluation score representation threshold value; if the evaluation score representation value exceeds the preset evaluation score representation threshold, carrying out numerical comparison on the interaction evaluation of the corresponding client and the preset interaction evaluation score threshold, and if the interaction evaluation score does not exceed the preset interaction evaluation score threshold, marking the corresponding client as a non-satisfied client;
acquiring the total online query time length and the online query times of the corresponding non-satisfied clients, and marking the corresponding non-satisfied clients as key reference clients if the total online query time length exceeds a preset online query time length threshold or the online query times exceed a preset query times threshold; performing numerical calculation on the number of unsatisfied clients and the number of key reference clients to obtain a client analysis value, and performing ratio calculation on the client analysis value and the number of referents to obtain a decision analysis value; comparing the decision analysis value with a preset decision analysis threshold value in a numerical value manner, and generating a investigation improvement signal if the decision analysis value exceeds the preset decision analysis threshold value; the investigation improvement signal is sent to a background monitoring end through a processor.
Furthermore, the invention also provides an intelligent customer service robot application method based on artificial intelligence, which comprises the following steps:
step one, receiving natural language query of a client, understanding query content of the client, and carrying out semantic analysis to identify the problem type of the client;
step two, matching the questions of the clients with the existing questions in the knowledge base module one by one, and when the questions of the clients are matched with the existing questions in the knowledge base module, calling corresponding answers from the knowledge base module and replying to the clients;
step three, when the problem matched with the problem of the customer does not exist in the knowledge base module, the intelligent reasoning reply module inquires related facts and information, deduces answers according to the related facts and information by applying a reasoning algorithm, and generates answers suitable for the customer according to the reasoning result and replies to the customer;
and fourthly, when the query content of the client cannot be identified or the answer is uncertain, realizing man-machine cooperation with the human customer service personnel through a man-machine interaction cooperation module.
Compared with the prior art, the invention has the beneficial effects that:
1. in the invention, the question types of the clients are identified by receiving natural language questions of the clients and understanding the query content of the clients and carrying out semantic analysis, the questions of the clients are matched with the existing questions in the knowledge base module one by one, when the questions of the clients are matched with the existing questions in the knowledge base module, corresponding answers are called from the knowledge base module and the clients are replied, when the questions matched with the questions of the clients do not exist in the knowledge base module, answers suitable for the clients are generated and the clients are replied through reasoning, and when the query content of the clients cannot be identified or the answers are uncertain, more accurate and humanized services are provided through man-machine cooperation; the problem distribution of the clients is judged by analyzing the problem distribution capturing and supervising module, the heat type and the frigidity type are determined, and the frigidity type is analyzed to judge whether the problem is of a failure type for answering, so that a manager can update and perfect a knowledge base of the corresponding type of problems in a targeted manner, and the use effect of the intelligent customer service robot is guaranteed;
2. in the invention, the operation performance evaluation module evaluates and analyzes the operation performance of the intelligent customer service robot to mark the corresponding detection time period as a high-quality operation time period or a low-quality operation time period, and generates an operation qualification signal or an operation disqualification signal through analysis so as to carry out retrospective investigation and make corresponding improvement measures in time for corresponding management staff, thereby further ensuring the use effect of the intelligent customer service robot; when the qualified evaluation signal is generated, the questionnaire sending and recycling module is used for issuing and recycling the questionnaire list, the consultable list is determined to improve the accuracy of the subsequent satisfaction decision analysis result, and the satisfaction decision feedback module is used for judging whether to generate the investigation improvement signal or not through analysis so as to perform function improvement of the intelligent robot in time for corresponding management staff, thereby ensuring the use effect and improving the customer satisfaction.
Drawings
For the convenience of those skilled in the art, the present invention will be further described with reference to the accompanying drawings;
FIG. 1 is a system block diagram of a first embodiment of the present invention;
FIG. 2 is a system block diagram of a second embodiment of the present invention;
fig. 3 is a flow chart of a method according to a third embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Embodiment one: as shown in FIG. 1, the intelligent customer service robot based on artificial intelligence provided by the invention comprises a processor, an inquiry receiving and processing module, a knowledge base module, an intelligent reasoning and replying module, a problem distribution capturing and supervising module and a running performance evaluation and supervising module; the query receiving processing module is used for receiving natural language query of the client, understanding query content of the client, carrying out semantic analysis to identify the problem type of the client, and sending the query content and the identification analysis information to the intelligent reasoning reply module through the processor; the knowledge base module stores common questions and answers of clients, the intelligent reasoning reply module matches the questions of the clients with the existing questions in the knowledge base module one by one, and when the questions of the clients are matched with the existing questions in the knowledge base module, the corresponding answers are called from the knowledge base module and the clients are replied;
when the problem matched with the problem of the client does not exist in the knowledge base module, the intelligent reasoning reply module inquires related facts and information, deduces answers by applying a reasoning algorithm according to the inquired related facts and information, generates answers suitable for the client according to the reasoning result and replies to the client; the reasoning algorithm comprises logic reasoning, probability reasoning and causal reasoning; also, logical reasoning can be divided into formal logic and non-formal logic. Formal logic refers to reasoning through a fixed reasoning mode, using formulas and rules. Non-formal logic refers to the fact that there is no fixed reasoning pattern, and reasoning needs to be performed through context, experience and intuition; the probability reasoning is to make reasoning by a probability theory method, and the probability value can represent the probability of occurrence of an event in the probability reasoning by considering the probability of occurrence of the event; causal reasoning refers to reasoning about causal relationships, i.e., the likelihood of extrapolating things from relationships between causes and results.
The intelligent reasoning reply module is in communication connection with the man-machine interaction cooperation module, and realizes man-machine cooperation with human customer service personnel through the man-machine interaction cooperation module when the inquiry content of the client cannot be identified or the answer is uncertain, so that more accurate and more humanized service is provided, and specific operations comprise intervention of the human customer service personnel, joint answer of the man-machine questions and supervision of the human customer service personnel; the human customer service personnel intervenes in the method and the system for automatically transferring the problems to the human customer service personnel when the intelligent customer service robot cannot identify or determine the inquiry content of the customer; the man-machine jointly answer questions are used for jointly answering the questions of the clients by the human customer service personnel and the intelligent customer service robot when the questions of the clients are complex and require more information and knowledge to answer, the human customer service personnel provides more information and knowledge, and the intelligent customer service robot provides standardized answers to improve the accuracy and humanization degree of the service; the human customer service personnel supervision is used for supervising and evaluating the answers of the intelligent customer service robot by the human customer service personnel, and when the answers of the robot are found to be wrong or inaccurate, the human customer service personnel can intervene in time and correct the answers.
The problem distribution capturing and supervising module is used for collecting all problem types of customer service, judging the problem distribution of the customer through analysis, determining the heat type and the frigidity type, analyzing the frigidity type to judge whether the problem type is a solution failure type, generating a perfect updating signal of the solution failure type and the heat type, transmitting the perfect updating signal to a background monitoring pipe end through the processor, and updating and perfecting a knowledge base of the solution failure type and the heat type by a manager of the background monitoring pipe end, thereby being beneficial to ensuring the use effect of the intelligent customer service robot; the specific operation process of the problem distribution capturing and supervising module is as follows:
collecting all problem types for customer service, marking the corresponding problem types as analysis objects i, i= {1,2, …, n }, wherein n represents the number of the problem types and n is a natural number greater than 1; acquiring an inquiry frequency value and an inquiry number quantity value of an analysis object i, wherein the inquiry frequency value is a data quantity value representing the number of inquires in unit time, and the inquiry number quantity value is a data quantity value representing the number of inquires in unit time; carrying out numerical calculation on the query frequency magnitude WP and the query number magnitude WR through a formula WX=qt1+qt2×WR to obtain a query coefficient WX; wherein qt1 and qt2 are preset weight coefficients, qt2 > qt1 > 0; comparing the query coefficient WX with a preset query coefficient threshold value in a numerical mode, if the query coefficient WX exceeds the preset query coefficient threshold value, indicating that the problem consultation of the corresponding analysis object i is more, marking the corresponding analysis object i as a hot type, and if the query coefficient WX is not in the preset query coefficient threshold value, indicating that the problem consultation of the corresponding analysis object i is less, marking the corresponding analysis object i as a cold type;
if the analysis object i is of a frigidity type, collecting client feedback information of the analysis object i, wherein the client feedback information comprises solution accuracy of each query, carrying out numerical comparison on the solution accuracy and a preset solution accuracy threshold, and if the solution accuracy does not exceed the preset solution accuracy threshold, marking a solution result of the corresponding query as a non-qualified solution result; calculating the ratio of the generation times of the unqualified answer result to the query times to obtain an unqualified answer value, and calculating the generation times FY of the unqualified answer result and the unqualified answer value FD through a formula DP=qw1 x FY+qw2 x FD to obtain an answer evaluation value DP; wherein qw1 and qw2 are preset weight coefficients, and qw2 > qw1 > 0; and, the larger the value of the answer evaluation value DP is, the worse the answer effect corresponding to the frigidity type is indicated; and (3) carrying out numerical comparison on the answer evaluation value DP and a preset answer evaluation threshold, and if the answer evaluation value DP exceeds the preset answer evaluation threshold, marking the corresponding frigidity type as an answer disqualification type.
The operation performance evaluation module evaluates and analyzes the operation performance of the intelligent customer service robot so as to mark the corresponding detection time period as a high-quality operation time period or a low-quality operation time period, generates an operation qualified signal or an operation unqualified signal through analysis, and sends the operation qualified signal or the operation unqualified signal to the processor; the processor sends the disqualification signal of the operation and assessment to a background monitoring end so as to timely carry out the traceability survey and make corresponding improvement measures corresponding to management staff, thereby further ensuring the use effect of the intelligent customer service robot; the specific operation process of the operation performance evaluation and supervision module is as follows:
the method comprises the steps that response speeds and processing speeds of an intelligent customer service robot for each inquiry in a detection period are collected, all the response speeds in the detection period are summed, an average value is obtained to obtain a response average speed, all the processing speeds in the detection period are summed, and the average value is obtained to obtain a processing average speed; performing numerical calculation on the response average speed XS and the processing average speed CS through a formula YX=et1×XS+et2×CS to obtain an initial analysis value YX; wherein, et1 and et2 are preset weight coefficients, and et1 is more than et2 is more than 0; and the value of the initial analysis value YX is in a direct proportion relation with the response average speed XS and the processing average speed CS, and the larger the value of the initial analysis value YX is, the faster the operation efficiency of the intelligent customer service robot corresponding to the detection period is;
collecting online inquiring people butted by the intelligent customer service robot in a detection period, presetting a plurality of online inquiring people ranges, and distributing a group of preset initial analysis threshold values for the online inquiring people ranges; comparing the online inquired number with all the online inquired number ranges one by one to determine a preset initial analysis threshold value of the fortune assessment corresponding to the detection period; it should be noted that, the larger the number value of the range of the number of the online inquires, the smaller the number value corresponding to the preset initial analysis threshold value of the fortune appraisal; comparing the initial analysis value YX with a corresponding preset initial analysis threshold value, and marking the corresponding detection period as a low-quality operation period if the initial analysis value YX does not exceed the corresponding preset initial analysis threshold value, which indicates that the operation efficiency of the intelligent customer service robot in the corresponding detection period is slower; if the initial analysis value YX of the operation evaluation exceeds the corresponding preset initial analysis threshold value of the operation evaluation, the operation efficiency of the intelligent customer service robot in the corresponding detection period is higher, and the corresponding detection period is marked as a high-quality operation period;
after marking the corresponding detection period as a low quality operation period or a high quality operation period, setting an operation evaluation period with a time length of K1, preferably, K1 is 36 hours; collecting the number of high-quality operation time periods and the number of low-quality operation time periods in an operation evaluation period, and calculating the ratio of the number of the low-quality operation time periods to the number of the high-quality operation time periods to obtain low-quality operation parameters; and comparing the low-quality operation parameter with a preset low-quality operation parameter threshold value, if the low-quality operation parameter does not exceed the preset low-quality operation parameter threshold value, indicating that the operation effect of the intelligent customer service robot in the operation evaluation period is good, generating an operation evaluation qualified signal, and if the low-quality operation parameter exceeds the preset low-quality operation parameter threshold value, indicating that the operation effect of the intelligent customer service robot in the operation evaluation period is poor, generating an operation evaluation unqualified signal.
Embodiment two: as shown in fig. 2, the difference between this embodiment and embodiment 1 is that the processor is communicatively connected with the questionnaire issuing and recovering module and the satisfaction decision feedback module, and when the processor receives the qualified evaluation signal, it sends the qualified evaluation signal to the questionnaire issuing and recovering module, and the questionnaire sending and recovering module generates a questionnaire table of the intelligent customer service robot, sends the questionnaire table to the served customer, and receives the questionnaire table fed back by the customer; the method comprises the steps of collecting a starting answering time and an ending answering time of a corresponding client, calculating a time difference between the ending answering time and the starting answering time to obtain an answering duration, comparing the answering duration with a preset answering duration in a numerical mode, if the answering duration does not exceed the preset answering duration, indicating that the feedback duration of the corresponding client is short, eliminating a questionnaire table of the corresponding client, marking the rest questionnaire table as a referent table, and remarkably improving the accuracy of a subsequent satisfaction decision analysis result;
all the referenceable tables are sent to a satisfaction decision feedback module through a processor; the specific operation process of the satisfaction decision feedback module is as follows: obtaining the evaluation results of each item in the corresponding referenceable table and assigning points to obtain interactive evaluation points of corresponding clients, summing all the interactive evaluation points and taking an average value to obtain an evaluation point representation value; comparing the evaluation score representation value with a preset evaluation score representation threshold value, and generating a investigation improvement signal if the evaluation score representation value does not exceed the preset evaluation score representation threshold value, which indicates that the satisfaction degree of the client to the intelligent customer service robot is lower; if the evaluation score representation value exceeds the preset evaluation score representation threshold, carrying out numerical comparison on the interaction evaluation of the corresponding client and the preset interaction evaluation score threshold, and if the interaction evaluation score does not exceed the preset interaction evaluation score threshold, indicating that the satisfaction of the corresponding client is poor, marking the corresponding client as a non-satisfied client;
acquiring the total online inquiring time length and the online inquiring times of the corresponding unsatisfactory clients, respectively carrying out numerical comparison on the total online inquiring time length and the online inquiring times with a preset total online inquiring time length threshold value and a preset online inquiring times threshold value, and marking the corresponding unsatisfactory clients as key reference clients if the total online inquiring time length exceeds the preset total online inquiring time length threshold value or the online inquiring times exceed the preset inquiring times threshold value; carrying out numerical calculation on the number FM of unsatisfactory clients and the number FZ of key reference clients through a formula KF= (bp 1. Times.FM+bp 2. Times.FZ)/2 to obtain a client analysis value KF; wherein, bp1 and bp2 are preset weight coefficients, and bp1 is more than bp2 and more than 1; calculating the ratio of the client analysis value to the number of the referenceable tables to obtain a decision analysis value;
it should be noted that, the larger the value of the decision analysis value is, the lower the satisfaction degree of the intelligent customer service robot is; comparing the decision analysis value with a preset decision analysis threshold value in a numerical value manner, and generating a investigation improvement signal if the decision analysis value exceeds the preset decision analysis threshold value; and the investigation improvement signal is sent to a background monitoring end through the processor so as to timely perform function improvement on the intelligent robot corresponding to the manager, thereby ensuring the use effect and improving the customer satisfaction.
Embodiment III: as shown in fig. 3, the difference between the present embodiment and embodiments 1 and 2 is that the application method of the intelligent customer service robot based on artificial intelligence provided by the present invention includes the following steps:
step one, receiving natural language query of a client, understanding query content of the client, and carrying out semantic analysis to identify the problem type of the client;
step two, matching the questions of the clients with the existing questions in the knowledge base module one by one, and when the questions of the clients are matched with the existing questions in the knowledge base module, calling corresponding answers from the knowledge base module and replying to the clients;
step three, when the problem matched with the problem of the customer does not exist in the knowledge base module, the intelligent reasoning reply module inquires related facts and information, deduces answers according to the related facts and information by applying a reasoning algorithm, and generates answers suitable for the customer according to the reasoning result and replies to the customer;
and fourthly, when the query content of the client cannot be identified or the answer is uncertain, realizing man-machine cooperation with the human customer service personnel through a man-machine interaction cooperation module.
The working principle of the invention is as follows: when the query receiving processing module is used, natural language query of a client is received, query content of the client is understood, semantic analysis is carried out, so that the problem type of the client is identified, the intelligent reasoning reply module matches the problems of the client with the problems existing in the knowledge base module one by one, and when the problems of the client are matched with the problems existing in the knowledge base module, corresponding answers are called from the knowledge base module and the client is replied; when the question matched with the question of the customer does not exist in the knowledge base module, generating an answer suitable for the customer through reasoning and replying to the customer, and when the query content of the customer cannot be identified or the answer is uncertain, realizing man-machine cooperation with human customer service personnel through a man-machine interaction cooperation module, so as to provide more accurate and humanized service; analyzing through a problem distribution capturing and supervising module to judge the problem distribution of a client, determining the heat type and the frigidity type, and analyzing the frigidity type to judge whether the problem is a solution failure type, so that a manager can update and perfect a knowledge base of the corresponding type of problem in a targeted manner, thereby being beneficial to ensuring the use effect of the intelligent customer service robot; the operation performance evaluation module evaluates and analyzes the operation performance of the intelligent customer service robot to mark the corresponding detection time period as a high-quality operation time period or a low-quality operation time period, and generates an operation evaluation qualified signal or an operation evaluation unqualified signal through analysis so as to timely trace and investigate the corresponding manager and make corresponding improvement measures, thereby further ensuring the use effect of the intelligent customer service robot.
The above formulas are all formulas with dimensions removed and numerical values calculated, the formulas are formulas with a large amount of data collected for software simulation to obtain the latest real situation, and preset parameters in the formulas are set by those skilled in the art according to the actual situation. The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof.
Claims (2)
1. The intelligent customer service robot based on the artificial intelligence is characterized by comprising a processor, an inquiry receiving and processing module, a knowledge base module, an intelligent reasoning and replying module, a problem distribution capturing and supervising module and a running performance evaluation and supervising module; the query receiving processing module is used for receiving natural language query of the client, understanding query content of the client, carrying out semantic analysis to identify the problem type of the client, and sending the query content and the identification analysis information to the intelligent reasoning reply module through the processor; the knowledge base module stores common questions and answers of clients, the intelligent reasoning reply module matches the questions of the clients with the existing questions in the knowledge base module one by one, and when the questions of the clients are matched with the existing questions in the knowledge base module, the corresponding answers are called from the knowledge base module and the clients are replied;
the problem distribution capturing and supervising module is used for collecting all problem types of customer service, judging the problem distribution of the customer through analysis, determining the heat type and the frigidity type, analyzing the frigidity type to judge whether the problem type is a solution failure type, generating a perfect updating signal of the solution failure type and the heat type, transmitting the perfect updating signal to a background monitoring pipe end through the processor, and updating and perfecting a knowledge base of the solution failure type and the heat type by a manager of the background monitoring pipe end; the operation performance evaluation module evaluates and analyzes the operation performance of the intelligent customer service robot so as to mark the corresponding detection time period as a high-quality operation time period or a low-quality operation time period, generates an operation qualified signal or an operation unqualified signal through analysis, and sends the operation qualified signal or the operation unqualified signal to the processor; the processor sends the unqualified evaluation signals to a background monitoring end;
when the problem matched with the problem of the client does not exist in the knowledge base module, the intelligent reasoning reply module inquires related facts and information, deduces answers by applying a reasoning algorithm according to the inquired related facts and information, generates answers suitable for the client according to the reasoning result and replies to the client; the reasoning algorithm comprises logic reasoning, probability reasoning and causal reasoning;
the intelligent reasoning reply module is in communication connection with the man-machine interaction cooperation module, and realizes man-machine cooperation with human customer service personnel through the man-machine interaction cooperation module when the inquiry content of the client cannot be identified or the answer is uncertain, so that more accurate and more humanized service is provided, and specific operations comprise intervention of the human customer service personnel, joint answer of the man-machine questions and supervision of the human customer service personnel;
the human customer service personnel intervenes in the method and the system for automatically transferring the problems to the human customer service personnel when the intelligent customer service robot cannot identify or determine the inquiry content of the customer; the man-machine jointly answer questions are used for jointly answering the questions of the clients by the human customer service personnel and the intelligent customer service robot when the questions of the clients are complex and require more information and knowledge to answer, the human customer service personnel provides more information and knowledge, and the intelligent customer service robot provides standardized answers to improve the accuracy and humanization degree of the service; human customer service personnel supervision is used for supervising and evaluating answers of the intelligent customer service robot by the human customer service personnel, and when the answers of the robot are found to be wrong or inaccurate, the human customer service personnel intervene in time and correct the answers;
the specific operation process of the problem distribution capturing and supervising module comprises the following steps:
collecting all problem types for customer service, marking the corresponding problem types as analysis objects i, i= {1,2, …, n }, wherein n represents the number of the problem types and n is a natural number greater than 1; acquiring the query frequency magnitude and the query number magnitude of the analysis object i, and carrying out numerical calculation on the query frequency magnitude and the query number magnitude to obtain a query coefficient; comparing the query coefficient with a preset query coefficient threshold value in a numerical mode, marking the corresponding analysis object i as a hot type if the query coefficient exceeds the preset query coefficient threshold value, and marking the corresponding analysis object i as a cold type if the query coefficient is not in the preset query coefficient threshold value;
if the analysis object i is of a frigidity type, collecting client feedback information of the analysis object i, wherein the client feedback information comprises solution accuracy of each query, carrying out numerical comparison on the solution accuracy and a preset solution accuracy threshold, and if the solution accuracy does not exceed the preset solution accuracy threshold, marking a solution result of the corresponding query as a non-qualified solution result; calculating the ratio of the generation times of the unqualified answer results to the query times to obtain unqualified answer values, and calculating the numerical value of the generation times of the unqualified answer results and the unqualified answer values to obtain answer evaluation values; the answer evaluation value is compared with a preset answer evaluation threshold value, and if the answer evaluation value exceeds the preset answer evaluation threshold value, the corresponding frigidity type is marked as an answer disqualification type;
the specific operation process of the operation performance evaluation and supervision module comprises the following steps:
the method comprises the steps that response speeds and processing speeds of an intelligent customer service robot for each inquiry in a detection period are collected, all the response speeds in the detection period are summed, an average value is obtained to obtain a response average speed, all the processing speeds in the detection period are summed, and the average value is obtained to obtain a processing average speed; carrying out numerical calculation on the response average speed and the processing average speed to obtain an initial analysis value of the operation and assessment;
collecting online inquiring people butted by the intelligent customer service robot in a detection period, presetting a plurality of online inquiring people ranges, and distributing a group of preset initial analysis threshold values for the online inquiring people ranges; comparing the online inquired number with all the online inquired number ranges one by one to determine a preset initial analysis threshold value of the fortune assessment corresponding to the detection period; comparing the initial analysis value with a corresponding preset initial analysis threshold value, and marking the corresponding detection period as a low-quality operation period if the initial analysis value does not exceed the preset initial analysis threshold value; otherwise, marking the corresponding detection time period as a high-quality operation time period;
after marking the corresponding detection time period as a low-quality operation time period or a high-quality operation time period, setting an operation evaluation period with the time length of K1, collecting the number of the high-quality operation time periods and the number of the low-quality operation time periods in the operation evaluation period, and calculating the ratio of the number of the low-quality operation time periods to the number of the high-quality operation time periods to obtain low-quality operation parameters; performing numerical comparison on the low-quality operation parameter and a preset low-quality operation parameter threshold, generating an operation evaluation qualified signal if the low-quality operation parameter does not exceed the preset low-quality operation parameter threshold, and generating an operation evaluation unqualified signal if the low-quality operation parameter exceeds the preset low-quality operation parameter threshold; and sending the qualified evaluation signal or unqualified evaluation signal to the processor;
the questionnaire sending and recycling module generates a questionnaire table of the intelligent customer service robot, sends the questionnaire table to a served customer, and receives the questionnaire table fed back by the customer; the method comprises the steps of collecting a starting answering time and an ending answering time of a corresponding client, calculating a time difference between the ending answering time and the starting answering time to obtain an answering duration, comparing the answering duration with a preset answering duration in a numerical mode, eliminating a questionnaire table of the corresponding client if the answering duration does not exceed the preset answering duration, marking the rest questionnaire table as a referenceable table, and sending all referenceable tables to a satisfaction decision feedback module through a processor;
the specific operation process of the satisfaction decision feedback module comprises the following steps:
obtaining the evaluation results of each item in the corresponding referenceable table and assigning points to obtain interactive evaluation points of corresponding clients, summing all the interactive evaluation points and taking an average value to obtain an evaluation point representation value; comparing the evaluation score representation value with a preset evaluation score representation threshold value in a numerical value manner, and generating a investigation improvement signal if the evaluation score representation value does not exceed the preset evaluation score representation threshold value; if the evaluation score representation value exceeds the preset evaluation score representation threshold, carrying out numerical comparison on the interaction evaluation of the corresponding client and the preset interaction evaluation score threshold, and if the interaction evaluation score does not exceed the preset interaction evaluation score threshold, marking the corresponding client as a non-satisfied client;
acquiring the total online query time length and the online query times of the corresponding non-satisfied clients, and marking the corresponding non-satisfied clients as key reference clients if the total online query time length exceeds a preset online query time length threshold or the online query times exceed a preset query times threshold; performing numerical calculation on the number of unsatisfied clients and the number of key reference clients to obtain a client analysis value, and performing ratio calculation on the client analysis value and the number of referents to obtain a decision analysis value; comparing the decision analysis value with a preset decision analysis threshold value in a numerical value manner, and generating a investigation improvement signal if the decision analysis value exceeds the preset decision analysis threshold value; the investigation improvement signal is sent to a background monitoring end through a processor.
2. An application method of an intelligent customer service robot based on artificial intelligence according to claim 1, comprising the following steps:
step one, receiving natural language query of a client, understanding query content of the client, and carrying out semantic analysis to identify the problem type of the client;
step two, matching the questions of the clients with the existing questions in the knowledge base module one by one, and when the questions of the clients are matched with the existing questions in the knowledge base module, calling corresponding answers from the knowledge base module and replying to the clients;
step three, when the problem matched with the problem of the customer does not exist in the knowledge base module, the intelligent reasoning reply module inquires related facts and information, deduces answers according to the related facts and information by applying a reasoning algorithm, and generates answers suitable for the customer according to the reasoning result and replies to the customer;
and fourthly, when the query content of the client cannot be identified or the answer is uncertain, realizing man-machine cooperation with the human customer service personnel through a man-machine interaction cooperation module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311208719.XA CN116955574B (en) | 2023-09-19 | 2023-09-19 | Intelligent customer service robot based on artificial intelligence and application method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311208719.XA CN116955574B (en) | 2023-09-19 | 2023-09-19 | Intelligent customer service robot based on artificial intelligence and application method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116955574A CN116955574A (en) | 2023-10-27 |
| CN116955574B true CN116955574B (en) | 2024-01-05 |
Family
ID=88442825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311208719.XA Active CN116955574B (en) | 2023-09-19 | 2023-09-19 | Intelligent customer service robot based on artificial intelligence and application method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116955574B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108256009A (en) * | 2018-01-03 | 2018-07-06 | 国网江苏省电力有限公司电力科学研究院 | A kind of method for improving electric intelligent response robot and answering accuracy rate |
| CN109325780A (en) * | 2018-08-24 | 2019-02-12 | 安徽讯飞智能科技有限公司 | A kind of exchange method of the intelligent customer service system in E-Governance Oriented field |
| CN110019149A (en) * | 2019-01-30 | 2019-07-16 | 阿里巴巴集团控股有限公司 | A kind of method for building up of service knowledge base, device and equipment |
| CN110232573A (en) * | 2018-03-06 | 2019-09-13 | 广州供电局有限公司 | Based on interactive intelligent response system |
| CN110893616A (en) * | 2018-09-13 | 2020-03-20 | 上海智臻智能网络科技股份有限公司 | Remote control method and device, computer equipment and storage medium |
| CN111913732A (en) * | 2020-08-28 | 2020-11-10 | 平安国际智慧城市科技股份有限公司 | Service updating method and device, management server and storage medium |
| KR20230044729A (en) * | 2021-09-27 | 2023-04-04 | 인포뱅크 주식회사 | An apparatus for chatbot making and diagnosis services and a method for operating it |
-
2023
- 2023-09-19 CN CN202311208719.XA patent/CN116955574B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108256009A (en) * | 2018-01-03 | 2018-07-06 | 国网江苏省电力有限公司电力科学研究院 | A kind of method for improving electric intelligent response robot and answering accuracy rate |
| CN110232573A (en) * | 2018-03-06 | 2019-09-13 | 广州供电局有限公司 | Based on interactive intelligent response system |
| CN109325780A (en) * | 2018-08-24 | 2019-02-12 | 安徽讯飞智能科技有限公司 | A kind of exchange method of the intelligent customer service system in E-Governance Oriented field |
| CN110893616A (en) * | 2018-09-13 | 2020-03-20 | 上海智臻智能网络科技股份有限公司 | Remote control method and device, computer equipment and storage medium |
| CN110019149A (en) * | 2019-01-30 | 2019-07-16 | 阿里巴巴集团控股有限公司 | A kind of method for building up of service knowledge base, device and equipment |
| CN111913732A (en) * | 2020-08-28 | 2020-11-10 | 平安国际智慧城市科技股份有限公司 | Service updating method and device, management server and storage medium |
| KR20230044729A (en) * | 2021-09-27 | 2023-04-04 | 인포뱅크 주식회사 | An apparatus for chatbot making and diagnosis services and a method for operating it |
Non-Patent Citations (1)
| Title |
|---|
| 人工智能在电子政务建设中的实例应用初探;李焱冬;;电子政务(第05期);87-93 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116955574A (en) | 2023-10-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109726200B (en) | Power grid information system fault positioning system and method based on bidirectional deep neural network | |
| CN106131158A (en) | Resource scheduling device based on cloud tenant's credit rating under a kind of cloud data center environment | |
| CN115599889B (en) | Intelligent answer method and system applied to online customer service of financial insurance platform | |
| US20230276000A1 (en) | Methods and internet of things systems for managing data of call centers of smart gas | |
| CN110309967A (en) | Prediction technique, system, equipment and the storage medium of customer service session grading system | |
| CN116028887B (en) | Analysis method of continuous industrial production data | |
| CN111666978A (en) | Intelligent fault early warning system for IT system operation and maintenance big data | |
| CN114936732A (en) | Informatization resource management system based on Internet | |
| CN116955574B (en) | Intelligent customer service robot based on artificial intelligence and application method thereof | |
| CN117094688B (en) | Digital control method and system for power supply station | |
| CN116911789A (en) | Borrowing monitoring management system suitable for tool and instrument use of manufacturing enterprises | |
| CN117592778A (en) | Risk early warning system | |
| WO2022151524A1 (en) | Method for seeking potential customers by using social software big data | |
| CN117172721A (en) | Data flow supervision early warning method and system for financing service | |
| CN112734447A (en) | Grain cloud after-market service cloud platform and method thereof | |
| CN110766248A (en) | Workshop human factor reliability evaluation method based on SHEL and interval intuition fuzzy evaluation | |
| CN115345710A (en) | E-commerce business big data decision analysis and data processing method | |
| CN115392832A (en) | Intelligent control system for plate processing | |
| CN113888231A (en) | Marketing data screening analysis system based on artificial intelligence | |
| CN113360628A (en) | Network repair event layered maintenance method and system based on problem knowledge base | |
| CN113449923A (en) | Multi-model object market quotation prediction method and device | |
| CN117556192B (en) | User growth type operation system and method based on artificial intelligence | |
| CN114912846B (en) | Comprehensive energy customer value mining energy efficiency evaluation method and device for online learning | |
| CN111242441B (en) | Self-adaptive parameter fitting method suitable for small micro-enterprise risk control model | |
| CN115277369A (en) | Base station alarm root cause accurate positioning method based on energy consumption data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |