CN116483979A - Dialog model training method, device, equipment and medium based on artificial intelligence - Google Patents
Dialog model training method, device, equipment and medium based on artificial intelligence Download PDFInfo
- Publication number
- CN116483979A CN116483979A CN202310573962.5A CN202310573962A CN116483979A CN 116483979 A CN116483979 A CN 116483979A CN 202310573962 A CN202310573962 A CN 202310573962A CN 116483979 A CN116483979 A CN 116483979A
- Authority
- CN
- China
- Prior art keywords
- answer
- text
- question
- initial
- sample
- 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 105
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 21
- 238000013519 translation Methods 0.000 claims abstract description 21
- 238000013145 classification model Methods 0.000 claims description 30
- 238000006243 chemical reaction Methods 0.000 claims description 22
- 238000004590 computer program Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 12
- 238000000605 extraction Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 11
- 238000011478 gradient descent method Methods 0.000 claims description 8
- 238000005516 engineering process Methods 0.000 abstract description 18
- 230000003321 amplification Effects 0.000 description 8
- 238000003199 nucleic acid amplification method Methods 0.000 description 8
- 230000003416 augmentation Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 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
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Human Computer Interaction (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Machine Translation (AREA)
Abstract
The invention is suitable for the field of financial science and technology, and particularly relates to a dialogue model training method, device, equipment and medium based on artificial intelligence. According to the invention, the initial question text, the initial answer text and the preset category prompt term are spliced into the first sample, the amplified question text, the amplified answer text and the amplified answer keyword are obtained through back-translation, and the second sample is obtained through splicing, so that the text data size is expanded, and the problem that the question text and the answer text are easy to be confused in the splicing result is solved; obtaining a first question-answer category of the initial answer keywords and splicing the first question-answer category into a label of a first sample, obtaining a second question-answer category of the amplified answer keywords and splicing the second question-answer category into a label of a second sample, and improving the reply performance of the dialogue model on different categories of questions by adding question-answer category information into the label; and the accuracy of the dialogue model is improved through training, and the dialogue accuracy of the customer service robot and the service efficiency and quality of financial business are improved in the field of financial science and technology.
Description
Technical Field
The invention is suitable for the field of financial science and technology, and particularly relates to a dialogue model training method, device, equipment and medium based on artificial intelligence.
Background
The conversation model can identify semantics according to information input by a user, then corresponding replies are generated according to the semantic information of the user, and along with the development of artificial intelligence technology, the utilization rate of the conversation model in virtual assistants, intelligent sound boxes and chatting conversations is gradually improved, for example, in the financial field, a virtual customer service robot can communicate with a customer based on the conversation model, and the conversation model has outstanding contribution in the aspects of solving customer questions, guiding customer transactions, providing after-sales services and the like, so that the service efficiency in the financial field is effectively improved.
The current dialogue model is mainly biased to directly splice dialogue questions and dialogue answers into input contents, and the dialogue model generates corresponding replies by extracting semantic features of the input contents.
Therefore, in the dialogue scenario in the financial field, how to improve the accuracy of the dialogue model is a problem to be solved.
Disclosure of Invention
In view of the above, the embodiments of the present invention provide a method, apparatus, device, and medium for training a dialogue model based on artificial intelligence, so as to solve the problem of low accuracy of the existing dialogue model.
In a first aspect, an embodiment of the present invention provides an artificial intelligence based dialog model training method, where the dialog model training method includes:
splicing the acquired initial question text, initial answer text and preset category prompt vocabulary terms, and determining a corresponding splicing result as a first sample;
extracting initial answer keywords according to the initial answer text, inputting the initial answer keywords into a pre-trained question-answer classification model to obtain a first question-answer category, splicing the initial answer keywords and the first question-answer category, and determining a splicing result as a label of the first sample;
respectively inputting the initial question text, the initial answer text and the initial answer keyword into a trained text back translation model to obtain an amplified question text, an amplified answer text and an amplified answer keyword, splicing the amplified question text, the amplified answer text and the preset category prompt term, and determining a corresponding splicing result as a second sample;
Inputting the amplified answer keywords into the pre-trained question-answer classification model to obtain a second question-answer category, splicing the amplified answer keywords and the second question-answer category, and determining that the splicing result is the label of the second sample;
and training a preset dialogue model by taking the first sample and the second sample as training samples and taking the label of the first sample and the label of the second sample as training labels to obtain a trained dialogue model.
In a second aspect, an embodiment of the present invention provides an artificial intelligence based dialog model training device, including:
the first sample splicing module is used for splicing the acquired initial question text, the initial answer text and the preset category prompt term, and determining a corresponding splicing result as a first sample;
the first label splicing module is used for extracting initial answer keywords according to the initial answer text, inputting the initial answer keywords into a pre-trained question-answer classification model to obtain a first question-answer category, splicing the initial answer keywords with the first question-answer category, and determining that a splicing result is a label of the first sample;
The second sample splicing module is used for respectively inputting the initial question text, the initial answer text and the initial answer keyword into a trained text back translation model to obtain an amplified question text, an amplified answer text and an amplified answer keyword, splicing the amplified question text, the amplified answer text and the preset category prompt term, and determining a corresponding splicing result as a second sample;
the second label splicing module is used for inputting the amplified answer keywords into the pre-trained question-answer classification model to obtain a second question-answer category, splicing the amplified answer keywords with the second question-answer category, and determining a splicing result as a label of the second sample;
and the dialogue model training module is used for training a preset dialogue model by taking the first sample and the second sample as training samples and taking the label of the first sample and the label of the second sample as training labels to obtain a trained dialogue model.
In a third aspect, an embodiment of the present invention provides a computer device, the computer device including a processor, a memory, and a computer program stored in the memory and executable on the processor, the processor implementing the dialog model training method as described in the first aspect when the computer program is executed.
In a fourth aspect, embodiments of the present invention provide a computer readable storage medium storing a computer program which, when executed by a processor, implements the dialog model training method as described in the first aspect.
Compared with the prior art, the embodiment of the invention has the beneficial effects that: the method comprises the steps of splicing an initial question text, an initial answer text and a preset category prompt term into a first sample, obtaining an amplified question text, an amplified answer text and an amplified answer keyword through a text back translation model, splicing the amplified question text, the amplified answer text and the preset category prompt term into a second sample, expanding text data, and distinguishing the question text and the answer text through setting the category prompt term, so that the problem that confusion is easy to occur in the question text and the answer text in a splicing result is solved; and the answer performance of the dialogue model on different types of questions is improved by adding the question-answer category information into the label by obtaining the first question-answer category of the initial answer keyword and splicing the first question-answer category into the label of the first sample and obtaining the second question-answer category of the amplified answer keyword and splicing the second question-answer category into the label of the second sample; and finally, the first sample and the second sample are used as training samples, and the label of the first sample and the label of the second sample are used as training labels to train a preset dialogue model, so that the accuracy of the trained dialogue model is effectively improved, the dialogue accuracy of the customer service robot in the financial science and technology field is improved, and the service efficiency and quality of financial business are further improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic view of an application environment of an artificial intelligence based dialog model training method according to an embodiment of the invention;
FIG. 2 is a flow chart of a dialog model training method based on artificial intelligence according to an embodiment of the present invention;
FIG. 3 is a schematic structural diagram of a dialogue model training device based on artificial intelligence according to a second embodiment of the present invention;
fig. 4 is a schematic structural diagram of a computer device according to a third embodiment of the present invention.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, techniques, etc., in order to provide a thorough understanding of the embodiments of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.
It should be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It should also be understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.
As used in the present description and the appended claims, the term "if" may be interpreted as "when..once" or "in response to a determination" or "in response to detection" depending on the context. Similarly, the phrase "if a determination" or "if a [ described condition or event ] is detected" may be interpreted in the context of meaning "upon determination" or "in response to determination" or "upon detection of a [ described condition or event ]" or "in response to detection of a [ described condition or event ]".
Furthermore, the terms "first," "second," "third," and the like in the description of the present specification and in the appended claims, are used for distinguishing between descriptions and not necessarily for indicating or implying a relative importance.
Reference in the specification to "one embodiment" or "some embodiments" or the like means that a particular feature, structure, or characteristic described in connection with the embodiment is included in one or more embodiments of the invention. Thus, appearances of the phrases "in one embodiment," "in some embodiments," "in other embodiments," and the like in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments" unless expressly specified otherwise. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless expressly specified otherwise.
The embodiment of the invention can acquire and process the related data based on the artificial intelligence technology. Among these, artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a digital computer-controlled machine to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use knowledge to obtain optimal results.
Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.
It should be understood that the sequence numbers of the steps in the following embodiments do not mean the order of execution, and the execution order of the processes should be determined by the functions and the internal logic, and should not be construed as limiting the implementation process of the embodiments of the present invention.
In order to illustrate the technical scheme of the invention, the following description is made by specific examples.
The dialog model training method based on artificial intelligence provided by the embodiment of the invention can be applied to an application environment as shown in fig. 1, wherein a client communicates with a server. The clients include, but are not limited to, palm top computers, desktop computers, notebook computers, ultra-mobile personal computer (UMPC), netbooks, cloud computing devices, personal digital assistants (personal digital assistant, PDA), and other computing devices. The server may be implemented by a stand-alone server or a server cluster formed by a plurality of servers.
The method can be widely applied to the fields of financial science and technology, internet science and technology, digital medical treatment, education and the like, for example, a customer service robot in the financial science and technology field can communicate with a customer according to a dialogue model trained by the dialogue model training method, so that the work of solving customer questions, guiding customer transactions and providing after-sales service is completed, and the service efficiency in the financial field is improved.
Referring to fig. 2, a flow chart of a dialogue model training method based on artificial intelligence according to an embodiment of the invention may be applied to the client in fig. 1, and the dialogue model training method may include the following steps:
step S201, the acquired initial question text, initial answer text and preset category prompt term are spliced, and a corresponding splicing result is determined to be a first sample.
When the dialogue model is trained, the initial question text and the initial answer text are used as the basis of the training text, and the question-answer characteristics are obtained by carrying out semantic analysis on the initial question text and the initial answer text. Correspondingly, aiming at the customer service robot in the financial field, the initial question text and the initial answer text can be questions and answers given by a customer in the process of communicating with the customer according to the manual customer service, so that a question and answer model corresponding to the customer service robot is trained based on the question and answer experience of the manual customer service, so that the question and answer model can give smooth and natural answers, and the accuracy of the question and answer model is improved.
The preset category prompting vocabulary term is a vocabulary term with category prompting information and is used for splicing with the initial question text and the initial answer text so as to simultaneously prompt the text category of the initial question text and the text category of the initial answer text. For example, the preset category prompting term may be "query" for prompting that the text category of the corresponding initial question text is "question" or "answer" for prompting that the text category of the corresponding initial answer text is "answer".
When the splicing result of the initial question text and the initial answer text is directly used as the input content of the dialogue model, the initial question text and the initial answer text in the splicing result are easy to be confused, and when the corresponding reply is generated by extracting the semantic features of the input content, the semantic understanding degree of the dialogue model on the input content is reduced, so that the accuracy of the dialogue model is reduced.
Therefore, in this embodiment, in order to improve accuracy of the dialogue model, the acquired initial question text, initial answer text and the preset category prompt term with text category prompt information are spliced, and the spliced result is used as a first sample in the training process of the dialogue model, so as to solve the problem that the initial question text and the initial answer text in the spliced result are easily confused, thereby improving semantic understanding degree of the dialogue model on input content, and further improving accuracy of the dialogue model.
Optionally, the splicing the obtained initial question text, the initial answer text and the preset category prompt term, and determining the corresponding splicing result as the first sample includes:
the preset category prompt vocabulary terms comprise a first sub vocabulary term and a second sub vocabulary term, and the acquired initial problem text and the first sub vocabulary term are spliced to obtain a problem prompt text;
splicing the obtained initial answer text and the second sub-term to obtain an answer prompt text;
and splicing the question prompt text and the answer prompt text, and determining a corresponding splicing result as a first sample.
The text content input by the model comprises two types of text, namely an initial question text and an initial answer text, and therefore a preset category prompting term is correspondingly set to comprise a first sub term and a second sub term, wherein the first sub term is used for prompting the text category of the corresponding initial question text, for example, the first sub term is 'query', and the second sub term is used for prompting the text category of the corresponding initial answer text, for example, the second sub term is 'answer'.
The method comprises the steps of splicing an obtained initial question text and a first sub-term to obtain a question prompt text, splicing an obtained initial answer text and a second sub-term to obtain an answer prompt text, splicing the question prompt text and the answer prompt text, and determining a corresponding splicing result as a first sample.
In an embodiment, the first sub-term is used as a prefix of the initial question text to be spliced with the initial question text, the second sub-term is used as a prefix of the initial answer text to be spliced with the initial answer text, and the initial question text and the initial answer text are separated while the text category of the initial question text and the text category of the initial answer text are prompted, so that the semantic understanding degree of the dialogue model on input content is further improved.
In another embodiment, the first sub-term is used as a suffix of the initial question text to be spliced with the initial question text, the second sub-term is used as a suffix of the initial answer text to be spliced with the initial answer text, and the initial question text and the initial answer text are separated while the text category of the initial question text and the text category of the initial answer text are prompted, so that the semantic understanding degree of the dialogue model on input content is further improved.
The step of splicing the acquired initial question text, the initial answer text and the preset category prompt term to determine the corresponding splicing result as the first sample solves the problem that the initial question text and the initial answer text are easy to be confused in the splicing result by splicing the initial question text, the initial answer text and the preset category prompt term with text category prompt information as the first sample in the dialogue model training process, improves the semantic understanding degree of the dialogue model on the input content, and further improves the accuracy of the dialogue model.
Step S202, extracting initial answer keywords according to the initial answer text, inputting the initial answer keywords into a pre-trained question-answer classification model to obtain a first question-answer category, splicing the initial answer keywords and the first question-answer category, and determining that the splicing result is a label of a first sample.
The initial answer keywords are answer keywords extracted from the initial answer text, and compared with the initial answer keywords, the initial answer text contains complex redundant information, and when the initial answer text is directly used as a label base of the dialogue model, the complex redundant information can reduce the accuracy of the dialogue model. Therefore, in the present embodiment, the initial answer keywords extracted from the initial answer text are used as the label base of the dialogue model, so as to improve the accuracy of the dialogue model.
Meanwhile, the extracted initial answer keywords are input into a pre-trained question-answer classification model to obtain a first question-answer category, the initial answer keywords and the first question-answer category are spliced, and the label of a first sample is determined as a splicing result. Wherein, the first question-answer category may be any one of address questions-answers, name questions-answers, time questions-answers, occupation questions-answers, and the like.
In one embodiment, a keyword extraction algorithm may be used to extract initial answer keywords from the initial answer text.
According to the method, the initial answer keywords are extracted according to the initial answer text, the initial answer keywords are input into the pre-trained question-answer classification model to obtain the first question-answer category, the initial answer keywords and the first question-answer category are spliced, the step of determining the splicing result as the label of the first sample is performed, the complex redundant information in the label is reduced by taking the initial answer keywords extracted from the initial answer text as the label basis of the dialogue model, the accuracy of the dialogue model is improved, and the answer performance of the dialogue model on different category questions is improved by adding the question-answer category information into the label.
Step S203, the initial question text, the initial answer text and the initial answer keywords are respectively input into a trained text back-translation model to obtain an amplified question text, an amplified answer text and an amplified answer keyword, the amplified question text, the amplified answer text and a preset category prompt term are spliced, and a corresponding splicing result is determined to be a second sample.
When the dialogue model is trained, the larger the data amount is, the better the dialogue reply effect of the dialogue model on different data is, so in the embodiment, in order to enable the dialogue model to learn the distribution characteristics of different data more fully so as to improve the model reply performance, the text reply model is used for carrying out language conversion on the initial question text, the initial answer text and the initial answer keywords to realize text amplification, and a richer text corpus is obtained.
Specifically, an initial question text, an initial answer text and an initial answer keyword are respectively input into a trained text back-translation model to obtain an amplified question text, an amplified answer text and an amplified answer keyword, the amplified question text, the amplified answer text and a preset category prompt term are further spliced, a corresponding splicing result is determined to be a second sample, the problem that the amplified question text and the amplified answer text are easy to be confused in the splicing result is solved, so that the semantic understanding degree of a dialogue model on input content is improved, complex redundant information in a label is reduced, and meanwhile question-answer category information is added into the label to improve the reply performance of the dialogue model on different category questions and improve the accuracy of the dialogue model.
Optionally, the initial question text, the initial answer text and the initial answer keyword are respectively input into a trained text back-translation model to obtain an amplified question text, an amplified answer text and an amplified answer keyword, including:
the text back translation model comprises a first language conversion sub-model and a second language conversion sub-model;
respectively inputting the initial question text, the initial answer text and the initial answer keyword into a first language conversion sub-model to obtain a first initial question text, a first initial answer text and a first initial answer keyword;
Respectively inputting the first initial question text, the first initial answer text and the first initial answer keyword into a second language conversion sub-model to obtain a second initial question text, a second initial answer text and a second initial answer keyword;
and performing de-duplication processing on the initial question text, the initial answer keyword, the second initial question text, the second initial answer text and the second initial answer keyword to obtain an amplified question text, an amplified answer text and an amplified answer keyword.
The text back-translation model comprises a first language conversion sub-model and a second language conversion sub-model, wherein the first language conversion sub-model is used for converting an initial question text, an initial answer text and an initial answer keyword from a class A language to a class B language, and the second language conversion sub-model is used for converting the initial question text, the initial answer text and the initial answer keyword converted to the class B language into the class A language again, namely, the back-translation is completed by carrying out language conversion on the initial question text, the initial answer text and the initial answer keyword for two times respectively, so as to obtain a second initial question text, a second initial answer text and a second initial answer keyword.
Since the text back-translation model performs language conversion on the initial question text, the initial answer text and the initial answer keyword to realize text expansion, there may be a case that the text difference before and after the language conversion is small, which causes redundancy of text data in the second sample and reduces training efficiency of the dialogue model.
Therefore, in order to improve the quality of the second sample, the embodiment further performs deduplication processing on the initial question text, the initial answer keyword, the second initial question text, the second initial answer text and the second initial answer keyword, so as to reduce the number of redundant texts obtained by back-translation, and obtain the amplified question text, the amplified answer text and the amplified answer keyword as construction bases of the second text.
Optionally, performing de-duplication processing on the initial question text, the initial answer keyword, the second initial question text, the second initial answer text and the second initial answer keyword to obtain an amplified question text, an amplified answer text and an amplified answer keyword, including:
calculating first similarity of the initial question text and the second initial question text, and taking the second initial question text with the first similarity not larger than a first preset threshold value as an amplified question text;
Calculating a second similarity between the initial answer text and the second initial answer text, and taking the second initial answer text with the second similarity not greater than a second preset threshold value as an amplified answer text;
and calculating a third similarity between the initial answer keywords and the second initial answer keywords, and taking the second initial answer keywords with the third similarity not larger than a third preset threshold value as amplified answer keywords.
The second initial question text, the second initial answer text and the second initial answer keyword are translated back based on the initial question text, the initial answer text and the initial answer keyword, and the embodiment performs duplicate removal processing by screening the second initial question text, the second initial answer text and the second initial answer keyword which have higher similarity with the initial question text, the initial answer text or the initial answer keyword.
Specifically, calculating first similarity of the initial question text and the second initial question text, and taking the second initial question text with the first similarity not larger than a first preset threshold value as an amplified question text; calculating a second similarity between the initial answer text and the second initial answer text, and taking the second initial answer text with the second similarity not greater than a second preset threshold value as an amplified answer text; and calculating a third similarity between the initial answer keywords and the second initial answer keywords, and using the second initial answer keywords with the third similarity not larger than a third preset threshold value as amplified answer keywords to finish the de-duplication processing of the initial question text, the initial answer keywords, the second initial question text, the second initial answer text and the second initial answer keywords.
The specific values of the first preset threshold, the second preset threshold and the third preset threshold can be set according to actual conditions.
According to the method, the initial question text, the initial answer text and the initial answer keywords are respectively input into the trained text back-translation model to obtain the amplified question text, the amplified answer text and the amplified answer keywords, the amplified question text, the amplified answer text and the preset category prompt terms are spliced into the second sample, the richer text corpus is obtained through back-translation, redundant texts are reduced through carrying out de-duplication processing on text data obtained through back-translation, the training efficiency of the dialogue model is improved, the amplified question text, the amplified answer text and the preset category prompt terms are spliced to obtain the second sample, the semantic understanding degree of the dialogue model on input contents is improved, and the accuracy of the dialogue model is further improved.
Step S204, the amplified answer keywords are input into a pre-trained question-answer classification model to obtain a second question-answer category, the amplified answer keywords and the second question-answer category are spliced, and the splice result is determined to be a label of a second sample.
The amplified answer keywords are input into the pre-trained question-answer classification model to obtain a second question-answer category, the amplified answer keywords and the second question-answer category are spliced, and a label with a second sample is determined as a splicing result. Wherein the second question-answer category may be any one of address questions-answers, name questions-answers, time questions-answers, occupation questions-answers, and the like.
Step S205, training a preset dialogue model by taking the first sample and the second sample as training samples and taking the label of the first sample and the label of the second sample as training labels to obtain a trained dialogue model.
The first sample and the second sample are the question text and the answer text which are amplified by the text and contain the text type prompt information, and the problem that the question text and the answer text are easy to be confused in a splicing result is effectively solved by splicing the preset type prompt term, so that the semantic understanding degree of a dialogue model on input content is improved; the labels of the first sample and the labels of the second sample are used for improving the reply performance of the dialogue model on different types of problems by adding question and answer type information into the labels.
Therefore, in this embodiment, the first sample and the second sample are used as training samples, and the label of the first sample and the label of the second sample are used as training labels, so that the preset dialogue model is trained, and the accuracy of the dialogue model is greatly improved.
Optionally, training a preset dialogue model by using the first sample and the second sample as training samples and using the label of the first sample and the label of the second sample as training labels to obtain a trained dialogue model, including:
The dialogue model comprises an encoder and a decoder, wherein a first sample and a second sample are used as training samples, and the training samples are input into the encoder for feature extraction to obtain question-answer features;
inputting the question-answer features into a decoder to obtain sample answers;
inputting the sample answers into a pre-trained question-answer classification model to obtain sample question-answer categories;
and calculating a first model loss according to the sample answers, the sample question-answer categories and the corresponding training labels, and reversely correcting parameters of the encoder and the decoder according to a gradient descent method until the first model loss converges to obtain a trained dialogue model.
The encoder is used for extracting features of the first sample and the second sample to obtain question-answer features, the decoder is used for decoding the question-answer features to output sample answers, and the training labels comprise question-answer category information, so that the output sample answers are input into a pre-trained question-answer classification model to obtain sample question-answer categories, and the first model loss can be calculated through the sample answers, the sample question-answer categories and the corresponding training labels to represent the difference degree between the output of the dialogue model and the labels.
The larger the first model loss, the smaller the accuracy of the dialogue model, and the parameters of the encoder and decoder need to be reversely corrected according to the gradient descent method until the first model loss converges to obtain a trained dialogue model.
Optionally, calculating the first model loss according to the sample answer, the sample question-answer category and the corresponding training label includes:
the training labels comprise corresponding initial answer keywords and corresponding question-answer categories;
calculating a first loss between the sample answers and the corresponding initial answer keywords, and calculating a second loss between the sample question-answer category and the first question-answer category or the second question-answer category;
substituting the first loss and the second loss into a preset loss relation model, and calculating to obtain the loss of the first model.
The training labels comprise corresponding initial answer keywords and corresponding question-answer categories, first losses between sample answers and the initial answer keywords are calculated respectively to represent differences between output sample answers and the initial answer keywords, and second losses between the sample question-answer categories and corresponding first question-answer categories or second question-answer categories are represented to represent differences between the obtained sample question-answer categories and the first question-answer categories or the second question-answer categories, and then the first losses and the second losses are substituted into a preset loss relation model to obtain first model losses through calculation.
For example, the number of sample answers, the number of initial answer keywords, the number of sample question-answer categories, the total number of first question-answer categories and second question-answer categories are all recorded as N, N is a positive integer, the i (i=1, 2, …, N) th sample answer is first converted into a sample answer vector according to the word vector technique, and recorded as X 1i Converting the ith initial answer key word into an initial answer key word vector and marking as X 2i The first penalty between the N sample answers and the N initial answer keywords is:
wherein L is 1 For the first loss, N is the number of sample answers and the number of initial answer keywords, X 1i For the i-th sample answer vector, X 2i I=1, 2, …, N for the i-th initial answer keyword vector.
When the sample question-answer category is consistent with the first question-answer category or the second question-answer category, the second loss is L 2 =0, when the sample question-answer category and the first question-answer category and the second question-answer category are not identical, the second loss is L 2 ＝1。
And substituting the first loss and the second loss into a preset loss relation model, and calculating to obtain the loss of the first model. In this embodiment, the predetermined loss relation model is a product of a first loss and a predetermined first loss weight, and the first model loss is:
L＝α 1 L 1 +α 2 L 2
wherein L is the first model loss, alpha 1 To preset the first loss weight, L 1 For the first loss, alpha 2 To preset the second loss weight, L 2 Is the second loss.
Wherein alpha is 1 And alpha 2 The specific value of (2) can be set by the actual situation, in the present embodiment, alpha is set according to the actual situation 1 ＝0.7，α 2 ＝0.3。
Optionally, the question-answer classification model comprises a classification encoder and a full-connection layer, wherein sample answer keywords are determined according to sample answers, the sample answer keywords are used as training samples, and the actual question-answer categories of the training samples are used as training labels;
the training process of the question-answer classification model comprises the following steps:
inputting the sample answer keywords into a classification encoder for feature extraction to obtain sample category features;
inputting sample category characteristics into a full-connection layer to obtain sample question-answer categories;
and calculating second model loss according to the sample question-answer category and the corresponding actual question-answer category, and reversely correcting parameters of the classification encoder and the full-connection layer according to the gradient descent method until the second model loss converges to obtain a trained question-answer classification model.
Firstly, extracting sample answer keywords in sample answers according to a keyword extraction algorithm, and training a question-answer classification model by taking the sample answer keywords as training samples and taking actual question-answer categories of the training samples as training labels.
Specifically, inputting sample answer keywords into a classification encoder for feature extraction to obtain sample category features, inputting the sample category features into a full-connection layer to obtain sample question-answer categories, calculating second model losses according to the sample question-answer categories and corresponding actual question-answer categories, and reversely correcting parameters of the classification encoder and the full-connection layer according to a gradient descent method until the second model losses converge to obtain a trained question-answer classification model if the second model losses are larger, which means that the classification accuracy of the question-answer classification model is lower.
According to the embodiment of the invention, the initial question text, the initial answer text and the preset category prompt term are spliced into the first sample, the amplified question text, the amplified answer text and the amplified answer keyword are obtained through the text back translation model, the amplified question text, the amplified answer text and the preset category prompt term are spliced into the second sample, the text data are expanded, the question text and the answer text are distinguished by setting the category prompt term, and the problem that the question text and the answer text in the spliced result are easy to be confused is solved; and the answer performance of the dialogue model on different types of questions is improved by adding the question-answer category information into the label by obtaining the first question-answer category of the initial answer keyword and splicing the first question-answer category into the label of the first sample and obtaining the second question-answer category of the amplified answer keyword and splicing the second question-answer category into the label of the second sample; and finally, the first sample and the second sample are used as training samples, and the label of the first sample and the label of the second sample are used as training labels to train a preset dialogue model, so that the accuracy of the trained dialogue model is effectively improved, the dialogue accuracy of the customer service robot in the financial science and technology field is improved, and the service efficiency and quality of financial business are further improved.
Corresponding to the dialogue model training method of the above embodiment, fig. 3 is a block diagram of a dialogue model training device based on artificial intelligence according to a second embodiment of the present invention, and for convenience of explanation, only the parts related to the embodiment of the present invention are shown.
Referring to fig. 3, the dialog model training device includes:
the first sample splicing module 31 is configured to splice the obtained initial question text, the initial answer text and the preset category prompt term, and determine that the corresponding splicing result is a first sample;
the first label splicing module 32 is configured to extract an initial answer keyword according to an initial answer text, input the initial answer keyword into a pre-trained question-answer classification model, obtain a first question-answer category, splice the initial answer keyword and the first question-answer category, and determine that a splice result is a label of a first sample;
the second sample splicing module 33 is configured to input an initial question text, an initial answer text and an initial answer keyword into the trained text back-translation model respectively, obtain an amplified question text, an amplified answer text and an amplified answer keyword, splice the amplified question text, the amplified answer text and a preset category prompt term, and determine that a corresponding splicing result is a second sample;
The second label splicing module 34 is configured to input the amplified answer keywords into the pre-trained question-answer classification model to obtain a second question-answer category, splice the amplified answer keywords and the second question-answer category, and determine that the splice result is a label of a second sample;
the dialogue model training module 35 is configured to train a preset dialogue model by using the first sample and the second sample as training samples and using the tag of the first sample and the tag of the second sample as training tags, so as to obtain a trained dialogue model.
Optionally, the first sample splicing module 31 includes:
the first splicing sub-module is used for presetting a category prompt term comprising a first sub-term and a second sub-term, and splicing the acquired initial problem text and the first sub-term to obtain a problem prompt text;
the second splicing sub-module is used for splicing the acquired initial answer text and the second sub-term to obtain an answer prompt text;
and the third splicing sub-module is used for splicing the question prompt text and the answer prompt text, and determining the corresponding splicing result as a first sample.
Optionally, the second sample splicing module 33 includes:
a text-back model determination sub-module for determining that the text-back model includes a first language-conversion sub-model and a second language-conversion sub-model;
The first language conversion sub-module is used for respectively inputting the initial question text, the initial answer text and the initial answer keyword into the first language conversion sub-model to obtain a first initial question text, a first initial answer text and a first initial answer keyword;
the second language conversion sub-module is used for respectively inputting the first initial question text, the first initial answer text and the first initial answer keyword into the second language conversion sub-module to obtain a second initial question text, a second initial answer text and a second initial answer keyword;
and the amplification data determination submodule is used for carrying out de-duplication processing on the initial question text, the initial answer keyword, the second initial question text, the second initial answer text and the second initial answer keyword to obtain an amplification question text, an amplification answer text and an amplification answer keyword.
Optionally, the amplification data determining submodule includes:
the first amplification data determining unit is used for calculating first similarity between the initial question text and the second initial question text, and taking the second initial question text with the first similarity not larger than a first preset threshold value as an amplification question text;
The second augmentation data determining unit is used for calculating a second similarity between the initial answer text and the second initial answer text, and taking the second initial answer text with the second similarity not larger than a second preset threshold value as an augmentation answer text;
and the third augmentation data determining unit is used for calculating a third similarity between the initial answer keywords and the second initial answer keywords, and taking the second initial answer keywords with the third similarity not larger than a third preset threshold value as augmentation answer keywords.
Optionally, the session model training module 35 includes:
the question-answering feature extraction submodule is used for a dialogue model comprising an encoder and a decoder, takes a first sample and a second sample as training samples, and inputs the training samples into the encoder for feature extraction to obtain question-answering features;
the question-answer feature decoding sub-module is used for inputting the question-answer features to the decoder to obtain sample answers;
the first question-answer classification sub-module is used for inputting sample answers into a pre-trained question-answer classification model to obtain sample question-answer categories;
and the first parameter correction sub-module is used for calculating a first model loss according to the sample answers, the sample question-answer categories and the corresponding training labels, and reversely correcting parameters of the encoder and the decoder according to the gradient descent method until the model loss converges to obtain a trained dialogue model.
Optionally, the first parameter correction submodule includes:
the training label determining unit is used for determining that the training label comprises a corresponding initial answer keyword and a corresponding question-answer category;
a first and second loss calculation unit for calculating a first loss between the sample answer and the corresponding initial answer keyword, and calculating a second loss between the sample question-answer category and the first question-answer category or the second question-answer category;
the model loss calculation unit is used for substituting the first loss and the second loss into a preset loss relation model to calculate and obtain the first model loss.
Optionally, the first question-answer classifying submodule includes:
a first question-answer classification sub-module determining unit for determining that the question-answer classification model includes a classification encoder and a full connection layer, determining sample answer keywords according to the sample answers, taking the sample answer keywords as training samples, and taking actual question-answer categories of the training samples as training labels;
the classification feature extraction unit is used for inputting the sample answer keywords into the classification encoder to perform feature extraction so as to obtain sample classification features;
the question-answer type determining unit is used for inputting sample type characteristics into the full-connection layer to obtain sample question-answer types;
And the second parameter correction unit is used for calculating second model loss according to the sample question-answer category and the corresponding actual question-answer category, and reversely correcting parameters of the classification encoder and the full-connection layer according to the gradient descent method until the second model loss converges to obtain a trained question-answer classification model.
It should be noted that, because the content of information interaction and execution process between the modules and the embodiment of the method of the present invention are based on the same concept, specific functions and technical effects thereof may be referred to in the method embodiment section, and details thereof are not repeated herein.
Fig. 4 is a schematic structural diagram of a computer device according to a third embodiment of the present invention. As shown in fig. 4, the computer device of this embodiment includes: at least one processor (only one shown in fig. 4), a memory, and a computer program stored in the memory and executable on the at least one processor, the processor executing the computer program to perform the steps of any of the various dialog model training method embodiments described above.
The computer device may include, but is not limited to, a processor, a memory. It will be appreciated by those skilled in the art that fig. 4 is merely an example of a computer device and is not intended to limit the computer device, and that a computer device may include more or fewer components than shown, or may combine certain components, or different components, such as may also include a network interface, a display screen, an input device, and the like.
The processor may be a CPU, but may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory includes a readable storage medium, an internal memory, etc., where the internal memory may be the memory of the computer device, the internal memory providing an environment for the execution of an operating system and computer-readable instructions in the readable storage medium. The readable storage medium may be a hard disk of a computer device, and in other embodiments may be an external storage device of the computer device, for example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), etc. that are provided on the computer device. Further, the memory may also include both internal storage units and external storage devices of the computer device. The memory is used to store an operating system, application programs, boot loader (BootLoader), data, and other programs such as program codes of computer programs, and the like. The memory may also be used to temporarily store data that has been output or is to be output.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, the specific names of the functional units and modules are only for distinguishing from each other, and are not used for limiting the protection scope of the present invention. The specific working process of the units and modules in the above device may refer to the corresponding process in the foregoing method embodiment, which is not described herein again. The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present invention may implement all or part of the flow of the method of the above-described embodiment, and may be implemented by a computer program to instruct related hardware, and the computer program may be stored in a computer readable storage medium, where the computer program, when executed by a processor, may implement the steps of the method embodiment described above. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, executable files or in some intermediate form, etc. The computer readable medium may include at least: any entity or device capable of carrying computer program code, a recording medium, a computer Memory, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), an electrical carrier signal, a telecommunications signal, and a software distribution medium. Such as a U-disk, removable hard disk, magnetic or optical disk, etc. In some jurisdictions, in accordance with legislation and patent practice, the computer readable medium may not be an electrical carrier signal or a telecommunications signal.
The present invention may also be implemented as a computer program product for implementing all or part of the steps of the method embodiments described above, when the computer program product is run on a computer device, causing the computer device to execute the steps of the method embodiments described above.
In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
In the embodiments provided by the present invention, it should be understood that the disclosed apparatus/computer device and method may be implemented in other manners. For example, the apparatus/computer device embodiments described above are merely illustrative, e.g., the division of modules or units is merely a logical functional division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection via interfaces, devices or units, which may be in electrical, mechanical or other forms.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the 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 scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention, and are intended to be included in the scope of the present invention.
Claims (10)
1. An artificial intelligence based dialog model training method, characterized in that the dialog model training method comprises:
splicing the acquired initial question text, initial answer text and preset category prompt vocabulary terms, and determining a corresponding splicing result as a first sample;
Extracting initial answer keywords according to the initial answer text, inputting the initial answer keywords into a pre-trained question-answer classification model to obtain a first question-answer category, splicing the initial answer keywords and the first question-answer category, and determining a splicing result as a label of the first sample;
respectively inputting the initial question text, the initial answer text and the initial answer keyword into a trained text back translation model to obtain an amplified question text, an amplified answer text and an amplified answer keyword, splicing the amplified question text, the amplified answer text and the preset category prompt term, and determining a corresponding splicing result as a second sample;
inputting the amplified answer keywords into the pre-trained question-answer classification model to obtain a second question-answer category, splicing the amplified answer keywords and the second question-answer category, and determining that the splicing result is the label of the second sample;
and training a preset dialogue model by taking the first sample and the second sample as training samples and taking the label of the first sample and the label of the second sample as training labels to obtain a trained dialogue model.
2. The method for training a dialogue model according to claim 1, wherein the splicing the acquired initial question text, initial answer text and preset category prompt term, and determining the corresponding splicing result as the first sample comprises:
the preset category prompt vocabulary terms comprise a first sub-vocabulary term and a second sub-vocabulary term, and the acquired initial problem text and the first sub-vocabulary terms are spliced to obtain a problem prompt text;
splicing the obtained initial answer text and the second sub-term to obtain an answer prompt text;
and splicing the question prompt text and the answer prompt text, and determining a corresponding splicing result as a first sample.
3. The dialog model training method of claim 1, wherein the inputting the initial question text, the initial answer text, and the initial answer keyword into the trained text back-translation model to obtain the amplified question text, the amplified answer text, and the amplified answer keyword includes:
the text back translation model comprises a first language conversion sub-model and a second language conversion sub-model;
respectively inputting the initial question text, the initial answer text and the initial answer keyword into the first language conversion sub-model to obtain a first initial question text, a first initial answer text and a first initial answer keyword;
Respectively inputting the first initial question text, the first initial answer text and the first initial answer keyword into the second language conversion sub-model to obtain a second initial question text, a second initial answer text and a second initial answer keyword;
and performing de-duplication processing on the initial question text, the initial answer keyword, the second initial question text, the second initial answer text and the second initial answer keyword to obtain an amplified question text, an amplified answer text and an amplified answer keyword.
4. The method for training a dialogue model according to claim 3, wherein performing a de-duplication process on the initial question text, the initial answer keyword, the second initial question text, the second initial answer text, and the second initial answer keyword to obtain an amplified question text, an amplified answer text, and an amplified answer keyword comprises:
calculating first similarity between the initial question text and the second initial question text, and taking the second initial question text with the first similarity not larger than a first preset threshold value as the amplified question text;
Calculating a second similarity between the initial answer text and the second initial answer text, and taking the second initial answer text with the second similarity not larger than a second preset threshold value as the amplified answer text;
and calculating a third similarity between the initial answer keywords and the second initial answer keywords, and taking the second initial answer keywords with the third similarity not larger than a third preset threshold value as the amplified answer keywords.
5. The method for training a conversation model according to claim 1, wherein the training a preset conversation model with the first sample and the second sample as training samples and with the label of the first sample and the label of the second sample as training labels to obtain a trained conversation model includes:
the dialogue model comprises an encoder and a decoder, the first sample and the second sample are used as training samples, and the training samples are input to the encoder for feature extraction to obtain question-answer features;
inputting the question-answer features to the decoder to obtain sample answers;
inputting the sample answers into a pre-trained question-answer classification model to obtain sample question-answer categories;
And calculating a first model loss according to the sample answers, the sample question-answer categories and the corresponding training labels, and reversely correcting parameters of the encoder and the decoder according to a gradient descent method until the first model loss converges to obtain a trained dialogue model.
6. The dialog model training method of claim 5, wherein the computing a first model loss from the sample answers, the sample question-answer categories, and corresponding training labels comprises:
the training label comprises a corresponding initial answer keyword and a corresponding question-answer category;
calculating a first loss between the sample answers and the corresponding initial answer keywords, and calculating a second loss between the sample question-answer category and the first question-answer category or the second question-answer category;
substituting the first loss and the second loss into a preset loss relation model, and calculating to obtain the loss of the first model.
7. The dialogue model training method as claimed in claim 5, wherein the question-answer classification model comprises a classification encoder and a full-connection layer, wherein sample answer keywords are determined according to the sample answers, the sample answer keywords are used as training samples, and actual question-answer categories of the training samples are used as training labels;
The training process of the question-answer classification model comprises the following steps:
inputting the sample answer keywords into the classification encoder for feature extraction to obtain sample category features;
inputting the sample category characteristics to the full-connection layer to obtain a sample question-answer category;
and calculating second model loss according to the sample question-answer category and the actual question-answer category, and reversely correcting parameters of the classification encoder and the full-connection layer according to a gradient descent method until the second model loss converges to obtain a trained question-answer classification model.
8. An artificial intelligence based dialog model training device, the dialog model training device comprising:
the first sample splicing module is used for splicing the acquired initial question text, the initial answer text and the preset category prompt term, and determining a corresponding splicing result as a first sample;
the first label splicing module is used for extracting initial answer keywords according to the initial answer text, inputting the initial answer keywords into a pre-trained question-answer classification model to obtain a first question-answer category, splicing the initial answer keywords with the first question-answer category, and determining that a splicing result is a label of the first sample;
The second sample splicing module is used for respectively inputting the initial question text, the initial answer text and the initial answer keyword into a trained text back translation model to obtain an amplified question text, an amplified answer text and an amplified answer keyword, splicing the amplified question text, the amplified answer text and the preset category prompt term, and determining a corresponding splicing result as a second sample;
the second label splicing module is used for inputting the amplified answer keywords into the pre-trained question-answer classification model to obtain a second question-answer category, splicing the amplified answer keywords with the second question-answer category, and determining a splicing result as a label of the second sample;
and the dialogue model training module is used for training a preset dialogue model by taking the first sample and the second sample as training samples and taking the label of the first sample and the label of the second sample as training labels to obtain a trained dialogue model.
9. A computer device, characterized in that it comprises a processor, a memory and a computer program stored in the memory and executable on the processor, which processor implements the dialog model training method according to any of claims 1 to 7 when the computer program is executed.
10. A computer readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the dialog model training method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310573962.5A CN116483979A (en) | 2023-05-19 | 2023-05-19 | Dialog model training method, device, equipment and medium based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310573962.5A CN116483979A (en) | 2023-05-19 | 2023-05-19 | Dialog model training method, device, equipment and medium based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116483979A true CN116483979A (en) | 2023-07-25 |
Family
ID=87216394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310573962.5A Pending CN116483979A (en) | 2023-05-19 | 2023-05-19 | Dialog model training method, device, equipment and medium based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116483979A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117216229A (en) * | 2023-11-08 | 2023-12-12 | 支付宝(杭州)信息技术有限公司 | Method and device for generating customer service answers |
-
2023
- 2023-05-19 CN CN202310573962.5A patent/CN116483979A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117216229A (en) * | 2023-11-08 | 2023-12-12 | 支付宝(杭州)信息技术有限公司 | Method and device for generating customer service answers |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113591902B (en) | Cross-modal understanding and generating method and device based on multi-modal pre-training model | |
| CN107273503B (en) | Method and device for generating parallel text in same language | |
| WO2021114840A1 (en) | Scoring method and apparatus based on semantic analysis, terminal device, and storage medium | |
| KR102316063B1 (en) | Method and apparatus for identifying key phrase in audio data, device and medium | |
| CN110569335B (en) | Triple verification method and device based on artificial intelligence and storage medium | |
| CN111611775B (en) | Entity identification model generation method, entity identification device and equipment | |
| CN110597966A (en) | Automatic question answering method and device | |
| CN113094478B (en) | Expression reply method, device, equipment and storage medium | |
| WO2022095370A1 (en) | Text matching method and apparatus, terminal device, and storage medium | |
| CN112951233A (en) | Voice question and answer method and device, electronic equipment and readable storage medium | |
| CN110532562B (en) | Neural network training method, idiom misuse detection method and device and electronic equipment | |
| CN115344699A (en) | Training method and device of text classification model, computer equipment and medium | |
| CN116483979A (en) | Dialog model training method, device, equipment and medium based on artificial intelligence | |
| CN111368066B (en) | Method, apparatus and computer readable storage medium for obtaining dialogue abstract | |
| CN112559711A (en) | Synonymous text prompting method and device and electronic equipment | |
| CN116680385A (en) | Dialogue question-answering method and device based on artificial intelligence, computer equipment and medium | |
| CN111401069A (en) | Intention recognition method and intention recognition device for conversation text and terminal | |
| CN116680387A (en) | Dialogue reply method, device, equipment and storage medium based on retrieval enhancement | |
| CN116186219A (en) | Man-machine dialogue interaction method, system and storage medium | |
| CN111753062A (en) | Method, device, equipment and medium for determining session response scheme | |
| CN113627186B (en) | Entity relation detection method based on artificial intelligence and related equipment | |
| CN113420545B (en) | Abstract generation method, device, equipment and storage medium | |
| CN113536790A (en) | Model training method and device based on natural language processing | |
| CN112925961A (en) | Intelligent question and answer method and device based on enterprise entity | |
| CN113421551B (en) | Speech recognition method, speech recognition device, computer readable medium and electronic equipment |
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 |