CN113821619A - Training method, device, system and computer readable storage medium - Google Patents
Training method, device, system and computer readable storage medium Download PDFInfo
- Publication number
- CN113821619A CN113821619A CN202111023604.4A CN202111023604A CN113821619A CN 113821619 A CN113821619 A CN 113821619A CN 202111023604 A CN202111023604 A CN 202111023604A CN 113821619 A CN113821619 A CN 113821619A
- Authority
- CN
- China
- Prior art keywords
- training
- user
- semantics
- answer
- model
- 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
- 238000012549 training Methods 0.000 title claims abstract description 172
- 238000000034 method Methods 0.000 title claims abstract description 59
- 238000011156 evaluation Methods 0.000 claims abstract description 53
- 238000004891 communication Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 4
- 238000013480 data collection Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- GNFTZDOKVXKIBK-UHFFFAOYSA-N 3-(2-methoxyethoxy)benzohydrazide Chemical compound COCCOC1=CC=CC(C(=O)NN)=C1 GNFTZDOKVXKIBK-UHFFFAOYSA-N 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
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
-
- 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
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3343—Query execution using phonetics
-
- 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
- G06F16/338—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
- G06Q50/2057—Career enhancement or continuing education service
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Data Mining & Analysis (AREA)
- Marketing (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- General Business, Economics & Management (AREA)
- Educational Technology (AREA)
- Databases & Information Systems (AREA)
- Tourism & Hospitality (AREA)
- Educational Administration (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Finance (AREA)
- Health & Medical Sciences (AREA)
- Accounting & Taxation (AREA)
- Acoustics & Sound (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Technology Law (AREA)
- Operations Research (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Primary Health Care (AREA)
- Mathematical Physics (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The invention discloses a training method, a device, a system and a computer readable storage medium, wherein the method comprises the following steps: when a starting instruction is received, displaying a service problem to a user, and receiving voice information sent by the user aiming at the service problem; comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain an evaluation result, and displaying the evaluation result and the answer of the business question to the user; according to the method and the device, the semantics of the answer of the business problem are compared with the semantics corresponding to the voice information sent by the user aiming at the business problem through the target model to obtain the evaluation result, and the evaluation result and the answer of the business problem are displayed to the user, so that the user can learn independently, and the training labor cost is saved.
Description
Technical Field
The invention relates to the technical field of machine learning, in particular to a training method, a training device, a training system and a computer readable storage medium.
Background
The business personnel who newly enter the insurance company often have deficient business knowledge, need spend a large amount of manpowers to train the business personnel who newly enter, but because the personnel of insurance trade just have high mobility by themselves, so will lead to being responsible for the training instructor and need training the business personnel who newly enter ceaselessly for the human cost of training the business personnel can not be released, caused the very big waste of human cost, therefore, how to practice thrift the human cost of training the business personnel, be the problem that the solution is badly needed.
Disclosure of Invention
The invention mainly aims to provide a training method, a training device, a training system and a computer readable storage medium, and aims to solve the problem of how to save the labor cost of training a salesman.
In order to achieve the above object, the present invention provides a training method, comprising the steps of:
when a starting instruction is received, displaying a service problem to a user, and receiving voice information sent by the user aiming at the service problem;
comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain an evaluation result, and displaying the evaluation result and the answer of the business question to the user.
Preferably, before the step of presenting the corresponding question to the user and receiving the voice message sent by the user for the question when the starting instruction is received, the training method further includes:
searching various service problems and corresponding answers through a data searching platform to obtain a training corpus set;
and training a target model based on the training corpus set.
Preferably, when receiving a start instruction, the step of presenting a service problem to a user and receiving voice information sent by the user for the service problem includes:
when a starting instruction is received, acquiring corresponding user information in the starting instruction, and determining a corresponding service problem set according to the user information;
and displaying the service problems in the service problem set to a user according to a preset sequence, and receiving voice information sent by the user aiming at the service problems.
Preferably, the target model includes a first model and a second model, and before the step of comparing the semantics of the answer to the business question with the semantics corresponding to the voice information through the target model, the training method further includes:
determining semantic judgment accuracy rates of the first model and the second model based on the training corpus set;
and comparing the semantic judgment accuracy of the first model with the semantic judgment accuracy of the second model to obtain a first comparison result, and determining the target model according to the first comparison result.
Preferably, the step of comparing the semantics of the answer to the business question with the semantics corresponding to the voice information through the target model to obtain the evaluation result includes:
comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain semantic similarity;
and comparing the semantic similarity with a preset similarity to obtain a second comparison result, and determining an evaluation result based on the second comparison result.
Preferably, the step of comparing the semantics of the answer to the service question with the semantics corresponding to the voice information through the target model to obtain the semantic similarity includes:
converting the semantics corresponding to the voice information into a first vector through a target model, and converting the semantics of the answer of the business question into a second vector;
and comparing the first vector with the second vector to obtain semantic similarity.
Preferably, after the step of comparing the semantics of the answer to the business question with the semantics corresponding to the voice information by using the object model to obtain an evaluation result, and presenting the evaluation result and the answer to the business question to the user, the training method further includes:
updating the training corpus set according to the voice information and the evaluation result, and executing the following steps: and training a target model based on the training corpus set.
In addition, to achieve the above object, the present invention provides a training apparatus, wherein the video generation apparatus includes:
the receiving module is used for displaying the service problem to a user and receiving the voice information sent by the user aiming at the service problem when receiving the starting instruction;
and the comparison module is used for comparing the semantics of the answer of the service question with the semantics corresponding to the voice information through the target model to obtain an evaluation result, and displaying the evaluation result and the answer of the service question to the user.
Further, the receiving module further comprises a training module, and the training module is configured to:
searching various service problems and corresponding answers through a data searching platform to obtain a training corpus set;
and training a target model based on the training corpus set.
Further, the receiving module is further configured to:
when a starting instruction is received, acquiring corresponding user information in the starting instruction, and determining a corresponding service problem set according to the user information;
and displaying the service problems in the service problem set to a user according to a preset sequence, and receiving voice information sent by the user aiming at the service problems.
Further, the comparison module is further configured to:
determining semantic judgment accuracy rates of the first model and the second model based on the training corpus set;
and comparing the semantic judgment accuracy of the first model with the semantic judgment accuracy of the second model to obtain a first comparison result, and determining the target model according to the first comparison result.
Further, the comparison module is further configured to:
comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain semantic similarity;
and comparing the semantic similarity with a preset similarity to obtain a second comparison result, and determining an evaluation result based on the second comparison result.
Further, the comparison module is further configured to:
converting the semantics corresponding to the voice information into a first vector through a target model, and converting the semantics of the answer of the business question into a second vector;
and comparing the first vector with the second vector to obtain semantic similarity.
Further, the comparison module further comprises an update module, and the update module is configured to:
updating the training corpus set according to the voice information and the evaluation result, and executing the following steps: and training a target model based on the training corpus set.
In addition, to achieve the above object, the present invention also provides a training system including: a memory, a processor, and a training program stored on the memory and executable on the processor, the training program when executed by the processor implementing the steps of the training method as described above.
In addition, to achieve the above object, the present invention also provides a computer readable storage medium, which stores a training program, and the training program, when executed by a processor, implements the steps of the training method as described above.
According to the training method, when a starting instruction is received, a business problem is displayed to a user, and voice information sent by the user aiming at the business problem is received; comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain an evaluation result, and displaying the evaluation result and the answer of the business question to the user; according to the method and the device, the semantics of the answer of the business problem are compared with the semantics corresponding to the voice information sent by the user aiming at the business problem through the target model to obtain the evaluation result, and the evaluation result and the answer of the business problem are displayed to the user, so that the user can learn independently, and the training labor cost is saved.
Drawings
FIG. 1 is a schematic diagram of an apparatus architecture of a hardware operating environment according to an embodiment of the present invention;
fig. 2 is a schematic flowchart of a training method according to a first embodiment of the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, fig. 1 is a schematic device structure diagram of a hardware operating environment according to an embodiment of the present invention.
The device of the embodiment of the invention can be a PC or a server device.
As shown in fig. 1, the apparatus may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration of the apparatus shown in fig. 1 is not intended to be limiting of the apparatus and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, the memory 1005, which is a type of computer storage medium, may include an operating system, a network communication module, a user interface module, and a training program.
The operating system is a program for managing and controlling the portable storage device and software resources, and supports the running of a network communication module, a user interface module, a training program and other programs or software; the network communication module is used for managing and controlling the network interface 1002; the user interface module is used to manage and control the user interface 1003.
In the storage device shown in fig. 1, the storage device calls a training program stored in a memory 1005 by a processor 1001 and performs operations in various embodiments of the training method described below.
Based on the hardware structure, the embodiment of the training method is provided.
Referring to fig. 2, fig. 2 is a schematic flow chart of a first embodiment of the training method of the present invention, which includes:
step S10, when receiving a starting instruction, displaying a service problem to a user, and receiving voice information sent by the user aiming at the service problem;
step S20, comparing the semantics of the answer of the service question with the semantics corresponding to the voice information through a target model to obtain an evaluation result, and displaying the evaluation result and the answer of the service question to the user.
The training method is applied to training equipment of insurance companies or other industries, the training equipment can be a PC, a mobile phone or a mobile terminal and the like, and has the function of receiving voice information; for convenience of description, the training equipment is taken as an example for description; when the training equipment receives a starting instruction, acquiring corresponding user information in the starting instruction, determining a corresponding business problem set according to the user information, displaying business problems in the business problem set to a user according to a preset sequence, and answering by the user according to the displayed business problems; when receiving a voice signal answered by a user aiming at a business question, the training equipment inputs the voice signal into a target model, the semantics of the answer of the business question are compared with the semantics corresponding to the voice information through the target model to obtain an evaluation result, the evaluation result is used for determining whether the answer of the user is correct or not, and the evaluation result and the answer of the business question are displayed to the user; it should be noted that, no matter whether the answer of the user to the business question is correct, the answer corresponding to the business question is displayed to the user, so that the user can know the answer corresponding to the business question, and a better training effect can be obtained.
According to the training method, when a starting instruction is received, the business problem is displayed to a user, and voice information sent by the user aiming at the business problem is received; comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through the target model to obtain an evaluation result, and displaying the answer of the business question to a user; according to the method and the device, the semantics of the answer of the business problem are compared with the semantics corresponding to the voice information sent by the user aiming at the business problem through the target model to obtain the evaluation result, and the answer of the business problem is displayed to the user, so that the user can learn independently, and the training labor cost is saved.
The respective steps will be described in detail below:
step S10, when receiving a starting instruction, displaying a service problem to a user, and receiving voice information sent by the user aiming at the service problem;
in the embodiment, when the training equipment receives the starting instruction, the user information in the starting instruction is obtained, the service problem set which needs to be trained of the user is determined according to the user information, the service problems in the service problem set are displayed to the user, the user answers according to the service problems displayed by the training equipment, and the training equipment receives the voice signals which are answered by the user aiming at the service problems.
Specifically, step S10 further includes:
step a, when a starting instruction is received, acquiring corresponding user information in the starting instruction, and determining a corresponding service problem set according to the user information;
in the step, when the training equipment receives a starting instruction, the corresponding user information in the starting instruction is obtained, and training required by a user is determined according to the user information so as to determine a corresponding business problem set; in one embodiment, a user can log in an account of the user through a mobile phone or a computer, so that a training device can receive a starting instruction, the starting instruction comprises user information, when the training device receives the starting instruction, the user information in the starting instruction is obtained, whether the user has a training record is determined through the user information, if the training record exists, the business problem is continuously displayed to the user according to the training record, if the training record does not exist, specific conditions such as departments, positions, engaged business contents and the like of the user are determined according to the user information, and then a business problem set corresponding to the user is determined; it should be noted that, the user may log in his own account through manual input or voice input; the training equipment has a recording function, can record the training record, the answer error record and the like of the user, and can continue to train the user according to the training record, the answer error record and the like so as to improve the mastery degree of the user on the business problems.
And b, displaying the service problems in the service problem set to a user according to a preset sequence, and receiving voice information sent by the user aiming at the service problems.
In the step, the training equipment displays the service problems in the service problem set to the user in sequence according to a preset sequence, when the user sees the corresponding service problems, the user speaks corresponding answers according to the service problems, and the training equipment receives voice information corresponding to the answers spoken by the user for the service problems; in one embodiment, after determining a service problem set corresponding to a user, a training device may determine a preset order for displaying service problems to the user according to the difficulty level or importance level of the service problems in the service problem set, optionally, the training device may preferentially display easier service problems to the user according to the difficulty level of the service problems, so that the user may answer the easier service problems first, optionally, the training device may preferentially display more important service problems to the user according to the importance level of the service problems, it may be understood that the more important service problems may be problems that the user often encounters when processing a specific service, or may be more basic service problems, the user may rapidly know the service by preferentially displaying the more important service problems, after the training device displays the service problems to the user, and starting a radio function, receiving voice information corresponding to an answer spoken by a user for the service question, determining the answer by pressing a determination button after the user finishes answering, or closing the radio function and determining the answer when the training equipment detects that the user does not answer any more. It should be noted that after the user finishes answering, if the user feels that there is a better answer, the user may choose to answer again.
Step S20, comparing the semantics of the answer of the service question with the semantics corresponding to the voice information through a target model to obtain an evaluation result, and displaying the evaluation result and the answer of the service question to the user.
In this embodiment, the training device inputs a voice signal sent by a user for a business question into the target model, compares the semantics of an answer corresponding to the business question with the semantics corresponding to the voice signal through the target model to obtain an evaluation result, determines whether the answer of the user is correct, and displays the evaluation result and the answer corresponding to the business question to the user in a text form. It should be noted that the target model includes a first model and a second model, where the first model is a simnet model, and the second model is a semantic representation model, including an ernie model and an gru model.
Specifically, step S20 is preceded by:
step c, determining the semantic judgment accuracy of the first model and the second model based on the training corpus set;
in the step, the training equipment respectively inputs the training corpora in the training corpus set into a first model and a second model, and then semantic judgment accuracy rates corresponding to the first model and the second model are respectively determined; in another embodiment, the training device averagely divides all the training corpora in the training corpus set into 10 parts according to specific settings, respectively inputs the 10 parts of training corpora into the first model and the second model, respectively records semantic prediction accuracy of the 10 parts of training corpora by the first model and the second model, and averages the 10 semantic prediction accuracy of the first model to obtain the semantic prediction accuracy corresponding to the first model, and averages the 10 semantic prediction accuracy of the second model to obtain the semantic prediction accuracy corresponding to the second model. It should be noted that the above process of determining the semantic prediction accuracy of the first model and the second model is only for illustration and not for limitation, and other feasible methods capable of determining the semantic prediction accuracy of the first model and the second model can be applied to this step, and are not limited herein.
And d, comparing the semantic judgment accuracy of the first model with the semantic judgment accuracy of the second model to obtain a first comparison result, and determining the target model according to the first comparison result.
In the step, the training equipment compares the semantic judgment accuracy of the first model with the semantic judgment accuracy of the second model to obtain a first comparison result, if the first comparison result indicates that the semantic judgment accuracy of the first model is greater than that of the second model, the target model is determined to be the first model, namely the simnet model, and if the first comparison result indicates that the semantic judgment accuracy of the first model is less than that of the second model, the target model is determined to be the second model, namely the semantic representation model, including the ernie model and the gru model. Specifically, step S20 includes:
step e, comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain semantic similarity;
in the step, the training equipment compares the semantics of the answer of the business question with the semantics corresponding to the voice information through the target model to obtain semantic similarity; it should be noted that, at the beginning of the operation of the training device, because the semantic representation model lacks the training corpus corresponding to the service, and the semantic judgment accuracy of the semantic representation model is lower than that of the first model, i.e., the simnet model, at the beginning of the operation of the training device, the first model, i.e., the simnet model, is basically used as the target model, and the semantic comparison is performed between the voice information of the user and the answer of the service question to obtain the semantic similarity; after the training equipment runs for a period of time, the semantic judgment accuracy of the semantic representation model is greater than that of the first model, namely the simnet model, at the moment, the second model, namely the ernie model and the gru model, is basically used as a target model, and the voice information of the user is semantically compared with the answer of the business question to obtain the semantic similarity.
Further, step e comprises:
converting the semantics corresponding to the voice information into a first vector through a target model, and converting the semantics of the answer of the business question into a second vector;
in the step, the training equipment converts the semantics corresponding to the voice information into a first vector through a target model and converts the semantics of the answer of the business question into a second vector; in one embodiment, when the target model is a first model, namely a simnet model, the training equipment inputs the obtained voice information and the answers to the business questions into the simnet model, converts the voice information into text information through the simnet model, converts the semantics corresponding to the text information into a first vector, and converts the semantics of the answers to the business questions into a second vector; in another embodiment, when the target model is the second model, namely the ernie model and the gru model, the training device inputs the obtained speech information and the answer to the business question into the ernie model, converts the speech information into text information through the ernie model, converts the semantic corresponding to the text information into the first vector, and converts the semantic of the answer to the business question into the second vector.
And comparing the first vector with the second vector to obtain semantic similarity.
In the step, the training equipment compares the first vector with the second vector to obtain semantic similarity; in one embodiment, when the target model is a first model, namely a simnet model, the training device compares the first vector with the second vector through the simnet model, and obtains semantic similarity between the voice information and the answer of the business question by calculating the distance between the first vector and the second vector; in another embodiment, when the target models are the second models, namely the ernie model and the gru model, the training device obtains the first vector and the second vector through the ernie model, then inputs the first vector and the second vector into the gru model for comparison, and calculates the distance between the first vector and the second vector through the gru model to obtain the semantic similarity between the speech information and the answer to the business question.
And f, comparing the semantic similarity with a preset similarity to obtain a second comparison result, and determining an evaluation result based on the second comparison result.
In the step, the training equipment compares the semantic similarity of the answer of the voice information and the service question with a preset similarity to obtain a second comparison result, if the semantic similarity of the answer of the voice information and the service question is greater than the preset similarity, the evaluation result is determined to be a correct answer of the user, and if the semantic similarity of the answer of the voice information and the service question is less than the preset similarity, the evaluation result is determined to be a wrong answer of the user; such as: the preset similarity preset in the training equipment by related research and development personnel is 70%, if the semantic similarity of the voice information obtained by the training equipment and the answer of the business question is more than 70%, the evaluation result is determined to be the correct answer of the user, and if the semantic similarity of the voice information obtained by the training equipment and the answer of the business question is less than 70%, the evaluation result is determined to be the wrong answer of the user.
Specifically, step S20 is followed by:
step g, updating the training corpus set according to the voice information and the evaluation result, and executing the steps of: and training a target model based on the training corpus set.
In the step, the training equipment updates the training corpus set based on the voice information and the evaluation result, trains the target model based on the updated training corpus set, and continues to perform other steps based on the trained target model; such as: the training corpus set comprises a voice information set with an assessment result of wrong answer of the user, a voice information set with a correct answer of the user and an answer set corresponding to a business question set, the training equipment adds the voice information with the wrong answer of the user to the voice information set with the wrong answer of the user or adds the voice information with the correct answer of the user to the voice information set with the correct answer of the user, the training corpus set is updated, and the target model is trained through the updated training corpus set, so that the semantic judgment accuracy of the target model is improved.
When the training equipment receives a starting instruction, acquiring corresponding user information in the starting instruction, determining a corresponding service problem set according to the user information, displaying service problems in the service problem set to a user according to a preset sequence, and answering by the user according to the displayed service problems; when the training equipment receives a voice signal answered by a user aiming at a business question, the voice signal is input into the target model, the semantics of the answer of the business question are compared with the semantics corresponding to the voice information through the target model to obtain an evaluation result, the evaluation result is used for determining whether the answer of the user is correct or not, and the evaluation result and the answer of the business question are displayed to the user, so that the user can learn independently, and the training labor cost is saved.
Further, based on the first embodiment of the training method of the present invention, a second embodiment of the training method of the present invention is provided.
The second embodiment of the training method differs from the first embodiment of the training method in that, before step S10, the training method further includes:
step h, searching various service problems and corresponding answers through a data searching platform to obtain a training corpus set;
and i, training a target model based on the training corpus set.
In this embodiment, the training device collects various service questions and corresponding answers through the data collection platform, where the answers corresponding to the various service questions include wrong answers and correct answers, and further obtain a corpus set, and trains the target model based on the corpus set, if: the method comprises the steps that related training personnel input various service problems and corresponding correct answers and wrong answers by logging in a data collection platform, training equipment obtains a training corpus set according to the various service problems and the corresponding correct answers and wrong answers, training is conducted on a target model based on the training corpus set, and it needs to be stated that the correct answers corresponding to the service literature comprise correct answers expressed by different semantics, for example, standard answers are 'account numbers and passwords are input firstly'. Then click to login ", other correct answers expressed by different semantics may be: such as: the method comprises the steps of inputting an account password in the first step, clicking a login button in the second step, inputting an account and inputting a password, clicking login, and the like.
The training device of the embodiment collects various service problems and corresponding answers through the data collection platform to obtain the training corpus set, and trains the target model based on the training corpus set, so that the target model meets the condition of online operation, and the training labor cost is saved.
The invention also provides a training device. The training device of the invention comprises:
the receiving module is used for displaying the service problem to a user and receiving the voice information sent by the user aiming at the service problem when receiving the starting instruction;
and the comparison module is used for comparing the semantics of the answer of the service question with the semantics corresponding to the voice information through the target model to obtain an evaluation result, and displaying the evaluation result and the answer of the service question to the user.
Further, the receiving module further comprises a training module, and the training module is configured to:
searching various service problems and corresponding answers through a data searching platform to obtain a training corpus set;
and training a target model based on the training corpus set.
Further, the receiving module is further configured to:
when a starting instruction is received, acquiring corresponding user information in the starting instruction, and determining a corresponding service problem set according to the user information;
and displaying the service problems in the service problem set to a user according to a preset sequence, and receiving voice information sent by the user aiming at the service problems.
Further, the comparison module is further configured to:
determining semantic judgment accuracy rates of the first model and the second model based on the training corpus set;
and comparing the semantic judgment accuracy of the first model with the semantic judgment accuracy of the second model to obtain a first comparison result, and determining the target model according to the first comparison result.
Further, the comparison module is further configured to:
comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain semantic similarity;
and comparing the semantic similarity with a preset similarity to obtain a second comparison result, and determining an evaluation result based on the second comparison result.
Further, the comparison module is further configured to:
converting the semantics corresponding to the voice information into a first vector through a target model, and converting the semantics of the answer of the business question into a second vector;
and comparing the first vector with the second vector to obtain semantic similarity.
Further, the comparison module further comprises an update module, and the update module is configured to:
updating the training corpus set according to the voice information and the evaluation result, and executing the following steps: and training a target model based on the training corpus set.
The invention also provides a training system.
The training system includes: a memory, a processor, and a training program stored on the memory and executable on the processor, the training program when executed by the processor implementing the steps of the training method as described above.
The method implemented when the training program running on the processor is executed may refer to various embodiments of the training method of the present invention, and details thereof are not repeated herein.
The invention also provides a computer readable storage medium.
The computer readable storage medium has stored thereon a training program which, when executed by the processor, implements the steps of the training method as described above.
The method implemented when the training program running on the processor is executed may refer to various embodiments of the training method of the present invention, and details thereof are not repeated herein.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention essentially or contributing to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (such as a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A training method, characterized in that the training method comprises the steps of:
when a starting instruction is received, displaying a service problem to a user, and receiving voice information sent by the user aiming at the service problem;
comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain an evaluation result, and displaying the evaluation result and the answer of the business question to the user.
2. The training method as claimed in claim 1, wherein before the step of presenting the corresponding question to the user upon receiving the start instruction and receiving the voice message issued by the user for the question, the training method further comprises:
searching various service problems and corresponding answers through a data searching platform to obtain a training corpus set;
and training a target model based on the training corpus set.
3. The training method as claimed in claim 1, wherein the step of presenting a business problem to a user upon receiving a start instruction and receiving voice information uttered by the user for the business problem comprises:
when a starting instruction is received, acquiring corresponding user information in the starting instruction, and determining a corresponding service problem set according to the user information;
and displaying the service problems in the service problem set to a user according to a preset sequence, and receiving voice information sent by the user aiming at the service problems.
4. A training method as claimed in claim 2, wherein the object model comprises a first model and a second model, and wherein before the step of comparing the semantics of the answer to the business question with the semantics of the speech information by the object model, the training method further comprises:
determining semantic judgment accuracy rates of the first model and the second model based on the training corpus set;
and comparing the semantic judgment accuracy of the first model with the semantic judgment accuracy of the second model to obtain a first comparison result, and determining the target model according to the first comparison result.
5. The training method as claimed in claim 4, wherein the step of comparing the semantics of the answer to the business question with the semantics of the speech information through the object model to obtain the evaluation result comprises:
comparing the semantics of the answer of the business question with the semantics corresponding to the voice information through a target model to obtain semantic similarity;
and comparing the semantic similarity with a preset similarity to obtain a second comparison result, and determining an evaluation result based on the second comparison result.
6. The training method as claimed in claim 5, wherein the step of comparing the semantics of the answer to the business question with the semantics corresponding to the voice information through the object model to obtain semantic similarity comprises:
converting the semantics corresponding to the voice information into a first vector through a target model, and converting the semantics of the answer of the business question into a second vector;
and comparing the first vector with the second vector to obtain semantic similarity.
7. The training method as claimed in claim 2, wherein after the step of comparing the semantics of the answer to the business question with the semantics of the speech information through the object model to obtain an evaluation result and presenting the evaluation result and the answer to the business question to the user, the training method further comprises:
updating the training corpus set according to the voice information and the evaluation result, and executing the following steps: and training a target model based on the training corpus set.
8. A training apparatus, characterized in that the training apparatus comprises:
the receiving module is used for displaying the service problem to a user and receiving the voice information sent by the user aiming at the service problem when receiving the starting instruction;
and the comparison module is used for comparing the semantics of the answer of the service question with the semantics corresponding to the voice information through the target model to obtain an evaluation result, and displaying the evaluation result and the answer of the service question to the user.
9. A training system, comprising: a memory, a processor, and a training program stored on the memory and executable on the processor, the training program when executed by the processor implementing the steps of the training method of any of claims 1 to 7.
10. A computer-readable storage medium, having stored thereon a training program, which when executed by a processor, performs the steps of the training method as recited in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111023604.4A CN113821619A (en) | 2021-08-31 | 2021-08-31 | Training method, device, system and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111023604.4A CN113821619A (en) | 2021-08-31 | 2021-08-31 | Training method, device, system and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113821619A true CN113821619A (en) | 2021-12-21 |
Family
ID=78923660
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111023604.4A Pending CN113821619A (en) | 2021-08-31 | 2021-08-31 | Training method, device, system and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113821619A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107274738A (en) * | 2017-06-23 | 2017-10-20 | 广东外语外贸大学 | Chinese-English translation teaching points-scoring system based on mobile Internet |
| CN110175229A (en) * | 2019-05-27 | 2019-08-27 | 言图科技有限公司 | A kind of method and system carrying out online training based on natural language |
| CN110458732A (en) * | 2019-06-17 | 2019-11-15 | 深圳追一科技有限公司 | Training Methodology, device, computer equipment and storage medium |
| CN110543557A (en) * | 2019-09-06 | 2019-12-06 | 北京工业大学 | construction method of medical intelligent question-answering system based on attention mechanism |
| CN112164262A (en) * | 2020-11-09 | 2021-01-01 | 河南环球优路教育科技有限公司 | Intelligent paper reading tutoring system |
| CN112328742A (en) * | 2020-11-03 | 2021-02-05 | 平安科技（深圳）有限公司 | Training method and device based on artificial intelligence, computer equipment and storage medium |
| CN112989826A (en) * | 2021-05-13 | 2021-06-18 | 平安科技（深圳）有限公司 | Test question score determining method, device, equipment and medium based on artificial intelligence |
-
2021
- 2021-08-31 CN CN202111023604.4A patent/CN113821619A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107274738A (en) * | 2017-06-23 | 2017-10-20 | 广东外语外贸大学 | Chinese-English translation teaching points-scoring system based on mobile Internet |
| CN110175229A (en) * | 2019-05-27 | 2019-08-27 | 言图科技有限公司 | A kind of method and system carrying out online training based on natural language |
| CN110458732A (en) * | 2019-06-17 | 2019-11-15 | 深圳追一科技有限公司 | Training Methodology, device, computer equipment and storage medium |
| CN110543557A (en) * | 2019-09-06 | 2019-12-06 | 北京工业大学 | construction method of medical intelligent question-answering system based on attention mechanism |
| CN112328742A (en) * | 2020-11-03 | 2021-02-05 | 平安科技（深圳）有限公司 | Training method and device based on artificial intelligence, computer equipment and storage medium |
| CN112164262A (en) * | 2020-11-09 | 2021-01-01 | 河南环球优路教育科技有限公司 | Intelligent paper reading tutoring system |
| CN112989826A (en) * | 2021-05-13 | 2021-06-18 | 平安科技（深圳）有限公司 | Test question score determining method, device, equipment and medium based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10217464B2 (en) | Vocabulary generation system | |
| CN111177413A (en) | Learning resource recommendation method and device and electronic equipment | |
| JP2017037588A (en) | Information processor and information processing program | |
| CN109155110B (en) | Information processing apparatus, control method therefor, and computer program | |
| CN116521841B (en) | Method, device, equipment and medium for generating reply information | |
| CN109543005A (en) | The dialogue state recognition methods of customer service robot and device, equipment, storage medium | |
| CN112417158A (en) | Training method, classification method, device and equipment of text data classification model | |
| CN108939532B (en) | Autism rehabilitation training guiding game type man-machine interaction system and method | |
| US20190116093A1 (en) | Simulating a user score from input objectives | |
| US20190114346A1 (en) | Optimizing user time and resources | |
| JP6442807B1 (en) | Dialog server, dialog method and dialog program | |
| CN111902831B (en) | Demonstration support system | |
| CN110647613A (en) | Courseware construction method, courseware construction device, courseware construction server and storage medium | |
| JP6030659B2 (en) | Mental health care support device, system, method and program | |
| JP2022189908A (en) | Point rating support device, and control method and control program therefor | |
| CN111444729A (en) | Information processing method, device, equipment and readable storage medium | |
| US20190197913A1 (en) | Information processing apparatus, information processing method, and non-transitory recording medium recorded with program | |
| CN111723559B (en) | Real-time information extraction method and device | |
| CN114186041A (en) | Answer output method | |
| CN113821619A (en) | Training method, device, system and computer readable storage medium | |
| CN113590772A (en) | Abnormal score detection method, device, equipment and computer readable storage medium | |
| CN117453887A (en) | Document-based question answering method and device, storage medium and computer equipment | |
| CN117556057A (en) | Knowledge question-answering method, vector database construction method and device | |
| CN116884282A (en) | Question answering method, device, electronic equipment and storage medium | |
| CN114841157A (en) | Online interaction method, system, equipment and storage medium based on data analysis |
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 |