CN115510216A - Customer service robot response method, system, computer equipment and storage medium - Google Patents
Customer service robot response method, system, computer equipment and storage medium Download PDFInfo
- Publication number
- CN115510216A CN115510216A CN202211325027.9A CN202211325027A CN115510216A CN 115510216 A CN115510216 A CN 115510216A CN 202211325027 A CN202211325027 A CN 202211325027A CN 115510216 A CN115510216 A CN 115510216A
- Authority
- CN
- China
- Prior art keywords
- client
- emotion
- processed
- information
- intention
- 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.)
- Pending
Links
Images
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
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention specifically discloses a customer service robot response method, which comprises the following steps: acquiring information to be processed sent by a client, and preprocessing and extracting basic features of the information to be processed; wherein the base features include intent features and emotional features; pre-classifying the information to be processed through a naive Bayes algorithm based on the extracted intention characteristics; performing emotion analysis through an emotion analysis model based on the extracted emotion characteristics; and executing corresponding system operation according to a first preset mechanism based on the pre-classification result, generating response data according to a second preset mechanism based on the emotion analysis result, and sending the result of the system operation and the response data to a client. The method has the advantages that the intention and the current emotion of a client can be quickly acquired, the requirement of the client can be timely and quickly processed, the emotion of the client can be taken care of, corresponding response measures can be provided, and the satisfaction degree of the client can be improved as much as possible.
Description
Technical Field
The invention belongs to the technical field of customer service robots, and particularly relates to a customer service robot response method, a customer service robot response system, computer equipment and a storage medium.
Background
With the prosperous development of the internet, the internet has wide spread range, strong interactivity and large amount of information, and the operation efficiency of the e-commerce enterprise is greatly improved, so that the scale of the e-commerce is gradually enlarged. The logistics industry has also been developed in great quantities as a supporting industry for electronic commerce. But at the same time, as the bill volume and data volume of the logistics industry are getting larger and larger, a plurality of problems which need to be solved urgently are brought. The online customer service and the automatic reply are an efficient application tool of the Internet in the logistics industry. By utilizing the online customer service and automatic reply functions, the client can check the running state of the logistics order by himself, the extra offline personnel and mechanisms are avoided, the enterprise operation cost is reduced, and meanwhile, the communication efficiency between the client and the client is improved. In the prior art, popular robot customer service generally can only perform good communication between express delivery with good logistics state and customers, but for some abnormal express deliveries, a good processing means cannot be provided for the requests of the customers, the emotion of the customers cannot be effectively identified, and the problem of low customer satisfaction is easily caused. Therefore, the information processing capacity of the customer service robot needs to be further improved urgently to meet the requirements of large business volume and many problems in the logistics industry.
Disclosure of Invention
In order to solve the above problems, an object of the present invention is to provide a customer service robot response method, system, computer device, and storage medium, which can effectively handle abnormal situations frequently occurring in the logistics process and take care of the customer emotion, so as to improve the abnormal handling capability of the customer service robot, and further achieve the purposes of replacing labor and reducing the operation cost of enterprises.
In order to realize the purpose, the technical scheme of the invention is as follows: a customer service robot response method comprises the following steps: acquiring information to be processed sent by a client, and preprocessing and extracting basic features of the information to be processed; wherein the base features include intent features and emotional features; pre-classifying the information to be processed through a naive Bayes algorithm based on the extracted intention characteristics; performing emotion analysis through an emotion analysis model based on the extracted emotion characteristics; and executing corresponding system operation according to a first preset mechanism based on the pre-classification result, generating response data according to a second preset mechanism based on the emotion analysis result, and sending the result of the system operation and the response data to a client.
In an embodiment of the present invention, the preprocessing the information to be processed and extracting the basic features further include: establishing a dictionary library, wherein the dictionary library comprises all words of a training corpus, each word corresponds to a unique identification number and is represented by a one-hot text; splitting a sentence into a plurality of words, comparing each word with the dictionary library one by one, if the word is in the dictionary library, successfully splitting the word, and if not, continuously splitting and matching until the word is successful; extracting the words successfully segmented as basic characteristics; wherein the dictionary repository includes an intention feature dictionary repository and an emotion feature dictionary repository, and the base features include intention features and emotion features.
In an embodiment of the present invention, the pre-classifying the information to be processed through a naive bayes algorithm based on the extracted intention features further comprises: constructing a classifier, wherein the characteristics of the classifier are characteristic words in an intention characteristic dictionary library, and the categories of the classifier comprise urging, address change, express interception and delay screening; wherein, the expression form of the classifier is as follows:
wherein P (c) k ) Is the prior probability of a class, P (x | c) k ) Is the class conditional probability of a sample with respect to a class, P (x) is the evidence factor used for normalization; acquiring historical information to be processed and intention information thereof, and training the classifier; inputting the extracted intention features into the classifier to obtain an intention category.
In an embodiment of the present invention, said executing the corresponding system operation according to the first preset mechanism based on the pre-classification result further comprises: under the condition that the pre-classified intention type is a catalysis, acquiring the invoice number corresponding to the client from an invoice number database, and sending the invoice number to a preset data form; under the condition that the pre-classified intention type is address change, acquiring a waybill number corresponding to the client from a waybill number database, acquiring a pre-modified address from the client, and updating the pre-modified address into a waybill target address database; under the condition that the pre-classified intention type is express interception, acquiring an invoice number corresponding to the client from an invoice number database, acquiring a target address corresponding to the invoice number from an invoice target address database, and updating the target address to a transit center of a local city where the single number target address is located; and under the condition that the pre-classified intention category is delay screening, acquiring all on-the-way waybill numbers corresponding to the client from the waybill number database, and screening out delayed orders according to preset rules.
In an embodiment of the present invention, the performing emotion analysis through an emotion analysis model based on the extracted emotion characteristics further includes: dividing a positive word list, a negative word list and a degree side word list based on the emotion feature words in the emotion feature dictionary library; extracting emotional characteristics in the information to be processed, and traversing the positive word list, the negative word list and the degree side word list; under the condition that the emotional features hit the corresponding word list/word list, processing corresponding weight according to a preset rule; and outputting the final weight value.
In an embodiment of the present invention, the generating response data according to a second preset mechanism based on the emotion analysis result further comprises: under the condition that the final weight value is within a first preset range, replying according to a first reply template; and replying according to a second reply template under the condition that the final weight value is within a second preset range.
In an embodiment of the present invention, the generating response data according to a second preset mechanism based on the emotion analysis result further includes: and upgrading the information to be processed to manual processing under the condition that the final weight value is within a third preset range.
Based on the same conception, the invention also provides a customer service robot response system, which is characterized by comprising the following components: the system comprises a characteristic extraction module, a characteristic extraction module and a characteristic analysis module, wherein the characteristic extraction module is used for acquiring information to be processed sent by a client, and preprocessing and basic characteristic extraction are carried out on the information to be processed; wherein the base features include intent features and emotional features; the pre-classification module is used for pre-classifying the information to be processed through a naive Bayesian algorithm based on the extracted intention characteristics; the emotion analysis module is used for carrying out emotion analysis through an emotion analysis model based on the extracted emotion characteristics; and the execution module is used for executing corresponding system operation according to a first preset mechanism based on the pre-classification result, generating response data according to a second preset mechanism based on the emotion analysis result, and sending the result of the system operation and the response data to the client.
Based on the same concept, the present invention also provides a computer apparatus comprising: a memory for storing a processing program; and the processor realizes the customer service robot response method when executing the processing program.
Based on the same concept, the invention also provides a readable storage medium, wherein the readable storage medium stores a processing program, and the processing program realizes the customer service robot response method when being executed by a processor.
After the technical scheme is adopted, compared with the prior art, the invention has the advantages that:
1. according to the invention, the intention characteristics of the to-be-processed information sent by the client are extracted through the model constructed by the naive Bayesian algorithm and the emotion characteristics are extracted through the emotion analysis model, so that the client service robot can rapidly acquire the intention and the current emotion of the client, can timely and rapidly process the requirement of the client under the condition of acquiring the intention of the client, and timely feeds back the client, thereby improving the service processing efficiency. By taking care of the emotion of the client and giving corresponding response measures, the satisfaction degree of the client can be improved as much as possible, and the dissatisfaction of the client is avoided.
2. The invention can make the dictionary database more perfect and more approximate to the information analysis capability of the manual customer service by continuously enriching and adjusting the words in the dictionary database in the operation process, thereby replacing the traditional manual customer service as much as possible and improving the service processing efficiency.
Drawings
The following detailed description of embodiments of the invention is provided in conjunction with the appended drawings, in which:
FIG. 1 is a flow chart of a customer service robot response method of the present invention;
FIG. 2 is a flow chart of the present invention for preprocessing the information to be processed and extracting the basic features;
FIG. 3 is a flow chart of the present invention for pre-classifying the information to be processed by a naive Bayes algorithm based on the extracted intention characteristics;
FIG. 4 is a flow chart of emotion analysis by an emotion analysis model based on the extracted emotional characteristics according to the present invention;
FIG. 5 is a flow chart illustrating the operation of the system according to the first predetermined mechanism based on the result of the pre-classification;
FIG. 6 is a flow chart of generating response data according to a second predetermined mechanism based on the emotion analysis result according to the present invention;
FIG. 7 is a schematic view of a customer service robot response system of the present invention;
FIG. 8 is a schematic diagram of an embodiment of a computer device of the present invention.
Detailed Description
The invention is described in further detail below with reference to the figures and specific examples. Advantages and features of the present invention will become apparent from the following description and from the claims. It is to be noted that the drawings are in a very simplified form and are each provided with a non-precise ratio for the purpose of facilitating and clearly facilitating the description of the embodiments of the present invention.
It should be noted that all the directional indicators (such as upper, lower, left, right, front, and rear … …) in the embodiment of the present invention are only used to explain the relative position relationship between the components, the motion situation, and the like in a specific posture (as shown in the drawing), and if the specific posture is changed, the directional indicator is changed accordingly.
Example one
According to the embodiment, the intention characteristics and the emotion characteristics of the information to be processed sent by the client are extracted, so that the customer service robot can quickly acquire the intention and the current emotion of the client, the demands of the client can be timely and quickly processed under the condition of acquiring the intention of the client, the client can be timely fed back, and the service processing efficiency is improved. By taking care of the emotion of the client and giving corresponding response measures, the satisfaction degree of the client can be improved as much as possible, and the dissatisfaction of the client is avoided.
Specifically, as shown in fig. 1, a flowchart of a first example of the customer service robot response method is shown. It comprises the following steps:
s100: acquiring information to be processed sent by a client, and preprocessing and extracting basic features of the information to be processed; wherein the base features include intent features and emotional features;
the intention characteristics and the emotion characteristics of the information to be processed sent by the client are extracted, so that the client service robot can quickly acquire the intention and the current emotion of the client, the client requirements can be timely and quickly processed under the condition of acquiring the intention of the client, the client can be timely fed back, and the service processing efficiency is improved. By taking care of the emotion of the client and giving corresponding response measures, the satisfaction degree of the client can be improved as much as possible, and the dissatisfaction of the client is avoided.
Wherein, the preprocessing the information to be processed and extracting the basic features further comprises:
s101: establishing a dictionary library, wherein the dictionary library comprises all words of a training corpus, each word corresponds to a unique identification number and is represented by a one-hot text;
s102: splitting a sentence into a plurality of words, comparing each word with the dictionary library one by one, if the word is in the dictionary library, successfully splitting the word, and if not, continuously splitting and matching until the word is successful;
s103: extracting the words successfully segmented as basic characteristics; wherein the dictionary repository includes an intention feature dictionary repository and an emotion feature dictionary repository, and the base features include intention features and emotion features.
For example, the intention feature dictionary library may include "where, a little faster, as soon as possible, a piece dispatch, an address change, an address error, a return, an interception … …", and the like, and when it is recognized that the words in the intention feature dictionary library exist in the information to be processed, the corresponding words are extracted as the intention features. Similarly, the emotional feature dictionary database may include "slow death, when, urgent, as soon as possible … …", and the like, and when the words are recognized to exist in the information to be processed, the corresponding words are extracted as the emotional features. Wherein, each word in the dictionary base needs to be set with a unique code to realize accurate matching.
The words in the dictionary database are enriched and adjusted continuously in the operation process, so that the dictionary database is more perfect and is closer to the information analysis capability of manual customer service, the traditional manual customer service can be replaced as far as possible, and the service processing efficiency is improved.
The customer service robot can automatically acquire the information to be processed sent by the customer through the QQ or the WeChat group.
After the intention characteristics of the information to be processed are acquired, the intention of the client needs to be effectively identified.
S200: pre-classifying the information to be processed through a naive Bayes algorithm based on the extracted intention characteristics;
the Bayesian classification algorithm is a general term of a large class of classification algorithms, and the probability that a sample possibly belongs to a certain class is used as a classification basis under the condition of conditional probability independence of the naive Bayesian classification algorithm.
Specifically, the pre-classifying the information to be processed through a naive bayes algorithm based on the extracted intention characteristics further comprises:
first, a classifier needs to be constructed.
S201: constructing a classifier, wherein the characteristics of the classifier are characteristic words in an intention characteristic dictionary library, and the category of the classifier comprises a urging partAddress change, express interception and delay screening; wherein, the expression form of the classifier is as follows:
wherein P (c) k ) Is the prior probability of a class, P (x | c) k ) Is the class conditional probability of a sample with respect to a class, P (x) is the evidence factor used for normalization;
the classifier is classified based on repeated work frequently encountered by logistics enterprises in the actual operation process, and the repetitive work is handed to the customer service robot for processing, so that the processing efficiency can be improved, and the operation cost can be reduced. Meanwhile, the categories of the classifiers can be correspondingly changed and increased along with the adjustment of the business function of the enterprise.
Secondly, historical data is required to be used for carrying out model training on the constructed classifier.
S202: acquiring historical information to be processed and intention information thereof, and training the classifier;
finally, after the model training is completed, the intention recognition can be carried out on the information to be processed based on the extracted intention characteristics.
S203: inputting the extracted intention features into the classifier to obtain an intention category.
The classifier based on the training completion can effectively identify the intention of information to be processed, and meanwhile, in the operation process, along with the continuous improvement of model training, the accuracy of model identification can be further improved.
Preferably, in the prior art, most of the customer service robots of the logistics enterprises can only recognize according to the keywords of the customer information, but the emotional care of the customers is almost not available. In the scheme of the embodiment, the customer service robot can correspondingly identify the emotion of the customer so as to provide a targeted conversation strategy.
S300: performing emotion analysis through an emotion analysis model based on the extracted emotion characteristics;
specifically, the performing of emotion analysis through an emotion analysis model based on the extracted emotional features further includes:
s301: dividing a positive word list, a negative word list and a degree side word list on the basis of the emotion feature words in the emotion feature dictionary library;
the purpose of this step is to build a dictionary library of emotion characteristics, which, like the above, can be enriched continuously during the operation.
S302: extracting emotional characteristics in the information to be processed, and traversing the positive word list, the negative word list and the degree side word list;
and calculating the emotion degree based on the extracted emotional characteristics.
S303: under the condition that the emotional features hit the corresponding word list/word list, processing corresponding weight according to a preset rule;
for example, when the emotional characteristics belong to a positive word list, the weight is +1, when the emotional characteristics belong to a negative word list, the weight is-1, and when the degree adverbs appear in the front of the positive word list, the negative word list and the negative word list, the weights are respectively +2, -2 and-2. The above is only an exemplary example, and specific numerical adjustments may also be made according to data in the operation process.
S304: and outputting the final weight value.
Based on the final weight value, a certain standard can be set, and reasonable threshold division is carried out on the emotion of the client.
S400: and executing corresponding system operation according to a first preset mechanism based on the pre-classification result, generating response data according to a second preset mechanism based on the emotion analysis result, and sending the result of the system operation and the response data to a client.
Further, the executing the corresponding system operation according to the first preset mechanism based on the pre-classification result further includes:
s401: under the condition that the pre-classified intention type is a catalysis, acquiring the invoice number corresponding to the client from an invoice number database, and sending the invoice number to a preset data form;
s402: under the condition that the pre-classified intention type is address change, acquiring an invoice number corresponding to the client from an invoice number database, acquiring a pre-modified address from the client, and updating the pre-modified address into an invoice target address database;
s403: under the condition that the pre-classified intention type is express interception, acquiring an invoice number corresponding to the client from an invoice number database, acquiring a target address corresponding to the invoice number from an invoice target address database, and updating the target address to a transit center of a local city where the single number target address is located;
s404: and under the condition that the pre-classified intention category is delay screening, acquiring all on-the-way waybill numbers corresponding to the client from the waybill number database, and screening out delayed orders according to preset rules.
In addition to the possible cases of S401 to S404 above, a reasonable classification setting may be performed specifically based on the case that can be defined as a conventional problem in the actual operation process of the logistics enterprise.
Further, the generating response data according to a second preset mechanism based on the emotion analysis result further comprises:
s501: under the condition that the final weight value is within a first preset range, replying according to a first reply template;
s502: and replying according to a second reply template under the condition that the final weight value is within a second preset range.
S503: and upgrading the information to be processed to manual processing under the condition that the final weight value is within a third preset range.
For example, when the emotion analysis model judges that the emotion of the client belongs to normal handling things according to the information sent by the client, only a regular reply is needed, for example, "XXX query/XXX logistics state query is about to be performed for you" or the like. In the case where the emotion analysis model judges that the customer has been in a state of emotional dissatisfaction based on the information sent from the customer, in addition to the normal system operation, the customer is required to send some soothing reply similar to "very sorry brings inconvenience to you, we treat you with firefly" or the like. And when the client is judged to be in an abnormal and unsatisfactory state, upgrading to a manual processing mode, and carrying out butt joint processing by a manual client. Through the hierarchical targeted reply and processing of the information, on one hand, the problems of more problems in the logistics industry and high manual customer service cost can be met as much as possible, on the other hand, the targeted processing can be performed on the relatively serious problems, the customer satisfaction is improved, and the problems that in the prior art, the robot customer service is mechanically replied, the processing capacity is low and customers are not full are avoided.
Example two
As shown in fig. 7, based on the same concept, the present invention further provides a customer service robot response system 600, where the customer service robot response system 600 includes: the feature extraction module 601 is configured to acquire information to be processed sent by a client, and perform preprocessing and basic feature extraction on the information to be processed; wherein the base features include intent features and emotional features; a pre-classification module 602, configured to pre-classify the to-be-processed information through a naive bayes algorithm based on the extracted intention features; an emotion analysis module 603 configured to perform emotion analysis through an emotion analysis model based on the extracted emotion features; the executing module 604 is configured to execute a corresponding system operation according to a first preset mechanism based on the pre-classification result, generate response data according to a second preset mechanism based on the emotion analysis result, and send a result of the system operation and the response data to the client.
EXAMPLE III
Based on the same concept, as shown in fig. 8, the present invention also provides a computer device 700, wherein the computer device 700 may have a relatively large difference due to different configurations or performances, and may include one or more processors 710 (CPUs) 710 (e.g., one or more processors) and a memory 720, one or more storage media 730 (e.g., one or more mass storage devices) storing an application 733 or data 732. Memory 720 and storage medium 730 may be, among other things, transient storage or persistent storage. The program stored in the storage medium 730 may include one or more modules (not shown), each of which may include a sequence of instructions operating on the computer device 700. Further, the processor 710 may be configured to communicate with the storage medium 730 to execute a series of instruction operations in the storage medium 730 on the computer device 700.
The computer device 700 may also include one or more power supplies 740, one or more wired or wireless network interfaces 750, one or more input-output interfaces 760, and/or one or more operating systems 731, such as Windows Server, mac OS X, unix, linux, freeBSD, and so forth. Those skilled in the art will appreciate that the computer device configuration illustrated in FIG. 8 does not constitute a limitation of the computer device and may include more or fewer components than illustrated, or some components may be combined, or a different arrangement of components.
The computer readable instructions, when executed by the processor, cause the processor to perform the steps of: a customer service robot response method comprises the following steps: acquiring information to be processed sent by a client, and preprocessing and extracting basic features of the information to be processed; wherein the base features include intent features and emotional features; pre-classifying the information to be processed through a naive Bayesian algorithm based on the extracted intention characteristics; performing emotion analysis through an emotion analysis model based on the extracted emotion characteristics; and executing corresponding system operation according to a first preset mechanism based on the pre-classification result, generating response data according to a second preset mechanism based on the emotion analysis result, and sending the system operation result and the response data to a client.
In an embodiment of the present invention, the preprocessing the information to be processed and extracting the basic features further include: establishing a dictionary library, wherein the dictionary library comprises all words of a training corpus, each word corresponds to a unique identification number and is represented by a one-hot text; splitting a sentence into a plurality of words, comparing each word with the dictionary library one by one, if the word is in the dictionary library, successfully splitting the word, and if not, continuously splitting and matching until the word is successful; extracting the words successfully segmented as basic characteristics; wherein the dictionary repository includes an intention feature dictionary repository and an emotion feature dictionary repository, and the base features include intention features and emotion features.
In an embodiment of the present invention, the pre-classifying the information to be processed by a naive bayes algorithm based on the extracted intention features further comprises: constructing a classifier, wherein the characteristics of the classifier are characteristic words in an intention characteristic dictionary library, and the categories of the classifier comprise urging, address change, express interception and delay screening; wherein, the expression form of the classifier is as follows:
wherein P (c) k ) Is the prior probability of a class, P (x | c) k ) Is the class conditional probability of the sample relative to the class, P (x) is the evidence factor used for normalization; acquiring historical information to be processed and intention information thereof, and training the classifier; inputting the extracted intention features into the classifier to obtain an intention category.
In an embodiment of the present invention, said executing the corresponding system operation according to the first preset mechanism based on the pre-classification result further comprises: under the condition that the pre-classified intention type is a catalysis, acquiring the invoice number corresponding to the client from an invoice number database, and sending the invoice number to a preset data form; under the condition that the pre-classified intention type is address change, acquiring an invoice number corresponding to the client from an invoice number database, acquiring a pre-modified address from the client, and updating the pre-modified address into an invoice target address database; under the condition that the pre-classified intention type is express interception, acquiring an invoice number corresponding to the client from an invoice number database, acquiring a target address corresponding to the invoice number from an invoice target address database, and updating the target address to a transit center of a local city where the single number target address is located; and under the condition that the pre-classified intention category is delay screening, acquiring all on-the-way waybill numbers corresponding to the client from the waybill number database, and screening out delayed orders according to preset rules.
In an embodiment of the present invention, the performing emotion analysis through an emotion analysis model based on the extracted emotional features further includes: dividing a positive word list, a negative word list and a degree side word list based on the emotion feature words in the emotion feature dictionary library; extracting emotional characteristics in the information to be processed, and traversing the positive word list, the negative word list and the degree side word list; under the condition that the emotional features hit the corresponding word list/word list, processing corresponding weight according to a preset rule; and outputting the final weight value.
In an embodiment of the present invention, the generating response data according to a second preset mechanism based on the emotion analysis result further comprises: under the condition that the final weight value is within a first preset range, replying according to a first reply template; and replying according to a second reply template under the condition that the final weight value is within a second preset range.
In an embodiment of the present invention, the generating response data according to a second preset mechanism based on the emotion analysis result further includes: and upgrading the information to be processed to manual processing under the condition that the final weight value is within a third preset range.
In one embodiment, a readable storage medium is provided, the computer readable instructions, when executed by one or more processors, cause the one or more processors to perform the above-described prepaid association processing method.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention, which is substantially or partly contributed by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A customer service robot response method is characterized by comprising the following steps:
acquiring information to be processed sent by a client, and preprocessing and extracting basic features of the information to be processed; wherein the base features include intent features and emotional features;
pre-classifying the information to be processed through a naive Bayes algorithm based on the extracted intention characteristics;
performing emotion analysis through an emotion analysis model based on the extracted emotion characteristics;
and executing corresponding system operation according to a first preset mechanism based on the pre-classification result, generating response data according to a second preset mechanism based on the emotion analysis result, and sending the result of the system operation and the response data to a client.
2. The customer service robot response method according to claim 1, wherein the preprocessing and the basic feature extraction of the information to be processed further comprise:
establishing a dictionary library, wherein the dictionary library comprises all words of a training corpus, each word corresponds to a unique identification number and is represented by a one-hot text;
splitting a sentence into a plurality of words, comparing each word with the dictionary library one by one, if the word is in the dictionary library, successfully splitting the word, and if not, continuously splitting and matching until the word is successful;
extracting the words successfully segmented as basic characteristics; wherein the dictionary repository includes an intention feature dictionary repository and an emotion feature dictionary repository, and the base features include intention features and emotion features.
3. The customer service robot response method of claim 2, wherein the pre-classifying the information to be processed based on the extracted intent features via a naive bayes algorithm further comprises:
constructing a classifier, wherein the characteristics of the classifier are characteristic words in an intention characteristic dictionary library, and the categories of the classifier comprise urging, address change, express interception and delay screening; wherein, the expression form of the classifier is as follows:
wherein P (c) k ) Is the prior probability of a class, P (x | c) k ) Is the class conditional probability of a sample with respect to a class, P (x) is the evidence factor used for normalization;
acquiring historical information to be processed and intention information thereof, and training the classifier;
inputting the extracted intention features into the classifier to obtain intention categories.
4. The customer service robot response method of claim 3, wherein the performing respective system operations according to a first preset mechanism based on the pre-classification result further comprises:
under the condition that the pre-classified intention type is a catalysis, acquiring the invoice number corresponding to the client from an invoice number database, and sending the invoice number to a preset data form;
under the condition that the pre-classified intention type is address change, acquiring an invoice number corresponding to the client from an invoice number database, acquiring a pre-modified address from the client, and updating the pre-modified address into an invoice target address database;
under the condition that the pre-classified intention type is express interception, acquiring an invoice number corresponding to the client from an invoice number database, acquiring a target address corresponding to the invoice number from an invoice target address database, and updating the target address to a transit center of a local city where the single number target address is located;
and under the condition that the pre-classified intention type is delay screening, acquiring all in-transit waybill numbers corresponding to the client from the waybill number database, and screening out delayed orders according to a preset rule.
5. The customer service robot response method of claim 2, wherein the performing emotion analysis by an emotion analysis model based on the extracted emotional features further comprises:
dividing a positive word list, a negative word list and a degree side word list on the basis of the emotion feature words in the emotion feature dictionary library;
extracting emotional characteristics in the information to be processed, and traversing the positive word list, the negative word list and the degree side word list;
under the condition that the emotional features hit the corresponding word list/word list, processing corresponding weight according to a preset rule;
and outputting the final weight value.
6. The customer service robot response method of claim 5, wherein the generating response data according to a second preset mechanism based on the emotion analysis result further comprises:
under the condition that the final weight value is within a first preset range, replying according to a first reply template;
and replying according to a second reply template under the condition that the final weight value is within a second preset range.
7. The customer service robot response method of claim 6, wherein the generating response data according to a second preset mechanism based on the emotion analysis result further comprises: and upgrading the information to be processed to manual processing under the condition that the final weight value is within a third preset range.
8. A customer service robot response system, comprising:
the system comprises a characteristic extraction module, a characteristic extraction module and a characteristic analysis module, wherein the characteristic extraction module is used for acquiring information to be processed sent by a client, and preprocessing and basic characteristic extraction are carried out on the information to be processed; wherein the base features include intent features and emotional features;
the pre-classification module is used for pre-classifying the information to be processed through a naive Bayesian algorithm based on the extracted intention characteristics;
the emotion analysis module is used for carrying out emotion analysis through an emotion analysis model based on the extracted emotion characteristics;
and the execution module is used for executing corresponding system operation according to a first preset mechanism based on the pre-classification result, generating response data according to a second preset mechanism based on the emotion analysis result, and sending the result of the system operation and the response data to the client.
9. A computer device, comprising:
a memory for storing a processing program;
a processor which, when executing the processing program, implements the customer service robot response method according to any one of claim 1 to claim 7.
10. A readable storage medium, characterized in that the readable storage medium has stored thereon a processing program, which when executed by a processor, implements the customer service robot response method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211325027.9A CN115510216A (en) | 2022-10-27 | 2022-10-27 | Customer service robot response method, system, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211325027.9A CN115510216A (en) | 2022-10-27 | 2022-10-27 | Customer service robot response method, system, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115510216A true CN115510216A (en) | 2022-12-23 |
Family
ID=84512394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211325027.9A Pending CN115510216A (en) | 2022-10-27 | 2022-10-27 | Customer service robot response method, system, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115510216A (en) |
-
2022
- 2022-10-27 CN CN202211325027.9A patent/CN115510216A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210103965A1 (en) | Account manager virtual assistant using machine learning techniques | |
| CN111340148B (en) | Training method of business classification model, business classification method and terminal | |
| CA2503427C (en) | Integrated task management systems and methods for executing rule-based operations | |
| CN104834651B (en) | Method and device for providing high-frequency question answers | |
| CN114119058B (en) | User portrait model construction method, device and storage medium | |
| US11423066B2 (en) | System and method for reducing user query ambiguity through chatbot clarifying questions | |
| US10977290B2 (en) | Transaction categorization system | |
| JP2020013521A (en) | Improved request handling | |
| CN114119057B (en) | User portrait model construction system | |
| US10922633B2 (en) | Utilizing econometric and machine learning models to maximize total returns for an entity | |
| US20210406771A1 (en) | Systems and methods for using artificial intelligence to evaluate lead development | |
| CN110956303A (en) | Information prediction method, device, terminal and readable storage medium | |
| JP7350206B1 (en) | Recruitment data management system and recruitment data management method | |
| CN115510216A (en) | Customer service robot response method, system, computer equipment and storage medium | |
| TWI752367B (en) | Intelligent dialogue management method and system based on natural language processing | |
| CN111680941A (en) | Premium recommendation method, device, equipment and storage medium | |
| CN111966805B (en) | Method, device, medium and electronic equipment for assisting in realizing session | |
| TWM590719U (en) | Intelligent conversation management system based on natural language processing | |
| Hristova et al. | RatingBot: A Text Mining Based Rating Approach. | |
| US20230351154A1 (en) | Automated processing of feedback data to identify real-time changes | |
| US20230350847A1 (en) | Graphical user interface for display of real-time feedback data changes | |
| US11914844B2 (en) | Automated processing and dynamic filtering of content for display | |
| CN114118817B (en) | Bank loan examination order dispatching method, device and system | |
| KR102540562B1 (en) | Method to analyze consultation data | |
| WO1993007575A1 (en) | Fuzzy retrieving device and its method, and membership function generator |
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 |