CN114860883A - Intelligent question and answer method and system - Google Patents
Intelligent question and answer method and system Download PDFInfo
- Publication number
- CN114860883A CN114860883A CN202210560246.9A CN202210560246A CN114860883A CN 114860883 A CN114860883 A CN 114860883A CN 202210560246 A CN202210560246 A CN 202210560246A CN 114860883 A CN114860883 A CN 114860883A
- Authority
- CN
- China
- Prior art keywords
- question
- data
- answer
- bits
- field
- 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
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000013135 deep learning Methods 0.000 claims abstract description 20
- 230000004044 response Effects 0.000 claims abstract description 8
- 238000011156 evaluation Methods 0.000 claims abstract description 7
- 238000005516 engineering process Methods 0.000 claims abstract description 6
- 238000012795 verification Methods 0.000 claims description 26
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 238000012549 training Methods 0.000 claims description 13
- 238000000605 extraction Methods 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 11
- 238000013210 evaluation model Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000007246 mechanism Effects 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000005065 mining Methods 0.000 abstract description 6
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 238000010801 machine learning Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000013519 translation Methods 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
- 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
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Computer Security & Cryptography (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Human Computer Interaction (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of artificial intelligence/Internet, and particularly provides an intelligent question-answering method, which comprises the following steps: s1, analyzing the role of the user; s2, deep analysis of user problems; s3, searching a corresponding feedback answer in the database based on the question; s4, response evaluation; and S5, deep learning of the database. Compared with the prior art, the method is based on the characteristics of a closed internal system, starts from user role mining, establishes a system question-answer model associated with the user role and a method, continuously trains a rear-end question-answer database on the basis of the model and on the basis of semantic recognition and deep learning technologies, and achieves quick response and higher answer precision to the user question-answer.
Description
Technical Field
The invention relates to the technical field of artificial intelligence/Internet, and particularly provides an intelligent question answering method and system.
Background
In the prior art, a research on the question and answer accuracy of an intelligent question and answer system is mainly embodied in that by establishing a model for identifying intentions, a question and answer intention of a user is firstly analyzed, and then classified responses are performed, for example, in patent CN 112632246A, a robot dialogue method, a device and a computer device based on deep learning: the application relates to artificial intelligence and provides a robot dialogue method and device based on deep learning and computer equipment. When an input operation based on the dialog box is detected, acquiring operation contents of a user in the dialog box, wherein the operation contents comprise input contents and/or deletion contents. And acquiring a history search log associated with the operation content, and generating information to be identified according to the operation content and the history search log. And according to the trained user intention recognition model, performing user intention recognition on the information to be recognized, determining a user intention recognition result corresponding to the operation content, and displaying the data item matched with the user intention recognition result. By adopting the method, the content input or deleted by the user in the dialog box is obtained in real time, the input information source is expanded, the intention of the client can be quickly and accurately identified, and the identification accuracy and the service quality of intelligent customer service or intelligent question and answer robots and the like in the application scene of user intention identification are improved.
For another example: patent CN 106294616A, an intelligent question-answering robot system based on mobile internet: the intelligent question-answering robot system based on the mobile internet comprises a user mobility unit, a question-answering robot control unit and a question-answering robot control unit, wherein the user mobility unit is used for realizing the tracking of the real-time position of a user, the description of the movement track of the user and the demarcation of the activity area of the user based on a mobile internet platform; the mobile internet and social network unit is used for determining that each user has different user attributes and user tasks, and the interaction between the users forms respective user relationship and social circles; the user interest point mining and personalized customization unit is used for mining the interest points of the user and pushing personalized information; the user query and knowledge base matching unit is used for mining and processing the text subjected to Chinese word segmentation, semantic analysis and syntactic analysis and sending the processed information to the intelligent question-answer database unit; and the intelligent question-answer database unit is used for updating and optimizing related contents in the question knowledge base and the answer knowledge base.
The above-mentioned technology is based on the user's question-answer intention or user's interest judgement, correspond to the question answer database answer to carry on screening, make up and correlate, realize the accurate answer to the user more. However, in real life, based on the difference of social identities of various groups of people, since professional terms are the same but have different meanings among various industries, the answers are often confused when the question-answering system is actually used.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a novel safety protection device which is reasonable in design, safe and applicable.
A further technical task of the present invention is to provide a solution that is highly practical.
The technical scheme adopted by the invention for solving the technical problems is as follows:
an intelligent question-answering method comprises the following steps:
s1, analyzing the role of the user;
s2, deep analysis of user problems;
s3, searching a corresponding feedback answer in the database based on the question;
s4, response evaluation;
and S5, deep learning of the database.
Further, in step S1, performing identity field verification on the user question to determine the role attribute of the user, where the identity field verification is performed in a questionnaire form or by real-name authentication;
the real-name authentication gateway is connected to a CA mechanism for identity authentication and returns an authentication result; the real-name authentication gateway is connected to a public security population base for identity verification and returns a verification result; and the real-name authentication gateway receives the verification result and the verification result.
Further, the data length of the identity field is 20 bits, the length of the USB Key serial number is 16 bits, a first formula is adopted according to the first 4 bits of data in the data of the identity field and the 16 bits of serial number to construct a 20-bit first Key of the Trivium algorithm, and the 20-bit first Key is adopted to obtain the data of the 16-bit output field in the Trivium algorithm;
the first formula is:
wherein uuidi is the first 4 bits of data of the identity field, serial number i-4 is the 16 bits of serial number of the USB Key, and Key1 is the 20 bits of first Key; and comparing the data of the 16-bit output field with the data of the 16 bits of the identity field except the first 4 bits, if the data of the 16-bit output field is the same as the data of the 16 bits of the identity field except the first 4 bits, the identity verification is successful, otherwise, the identity verification fails.
Further, according to the 16-bit data of the identity field divided by the first 4 bits and the first 4-bit data of the first field, a second formula is adopted to construct a 20-bit second key of the Trivium algorithm, and the 20-bit second key is adopted to decrypt the data of the first field divided by the first 4 bits so as to obtain an actual value of the data of the first field divided by the first 4 bits;
the second formula is:
wherein uuidi-4 is 16 bits of data except the first 4 bits of the identity field, data i is the first 4 bits of data of the first field, and key2 is the 20 bits of second key.
Further, in step S2, fuzzy answers and accurate answers are set on the search interface, a dialogue record is obtained, the dialogue content is converted into a text form, a question and answer document is constructed, operations of segmenting words, removing stop words and assigning id numbers are performed on the question and answer document, so as to obtain a data set, the data set is used as an input of the deep learning network model, and the deep learning network model is trained for multiple times to form a question and answer system model.
Further, in step S3, when an answer request carrying a target question is received, semantic proximity between each answer and the target question is determined according to the target question, each answer in the answer query library, the question semantic extraction formula, the answer semantic extraction formula, the training value of the question semantic extraction parameter, and the training value of the answer semantic extraction parameter; and selecting the target answers from the answers according to the semantic proximity of the answers to the target question and the text proximity of the answers to the target question, and feeding back the answer request.
Further, in step S4, an evaluation model is constructed, and the evaluation model obtains a plurality of candidate answers from the answer text according to the query text, and evaluates the candidate answers through the convex neural network; and obtaining the optimal answer in the multiple candidate answers according to the evaluation result.
Further, in step S5, based on step S4, performing question-answer database model training, and refining and adjusting question-answer vocabulary in the question-answer database through a deep learning technique;
and analyzing the characteristic information of the target initialization neural network in preset neural network big data or a database to obtain an analysis result.
Further, the model structure of the question-answer database model is divided into three layers:
the method comprises an input layer, a hidden layer and a softmax layer, wherein onehot vectors are input into the input layer, a weight matrix w is arranged between the input layer and the hidden layer, the hidden layer is set to be 300-dimensional features, the weight matrix w and the onehot vectors are multiplied to obtain a 300-dimensional feature vector, and then probability output is obtained through a wx + b function and then softmax.
An intelligent question-answering system comprising: at least one memory and at least one processor;
the at least one memory to store a machine readable program;
the at least one processor is used for calling the machine readable program and executing an intelligent question answering method.
Compared with the prior art, the intelligent question answering method and the intelligent question answering system have the following outstanding beneficial effects:
the invention is based on the characteristics of a closed internal system, starts from user role mining, establishes a system question-answer model and a method which are associated with the user role, and continuously trains a rear-end question-answer database on the basis of the model and on the basis of semantic recognition and deep learning technologies, thereby realizing quick response and higher answer precision of the user question-answer.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of an intelligent question answering method.
Detailed Description
The present invention will be described in further detail with reference to specific embodiments in order to better understand the technical solutions of the present invention. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
A preferred embodiment is given below:
as shown in fig. 1, in the intelligent question-answering method in this embodiment, a user role is analyzed and confirmed, a user question is analyzed, a corresponding feedback answer is searched for based on the user question in a database corresponding to the user role, a system question-answering model and a method associated with the user role are established based on user role mining, a back-end question-answering database is continuously trained based on the model and based on semantic recognition and deep learning technologies, fast response to the user question-answering and higher answer accuracy are achieved, and confusion of answers is avoided.
First, the user role in this document refers to the identity that the user plays in a social environment in real life, for example: students, teachers, interior designers, architectural designers, etc., as well as parents, couples, pregnant women, etc.
The method comprises the following specific steps:
s1, user role analysis:
and carrying out identity field verification on the user question to determine the role attribute of the user, namely confirming the identity attribute played by the user in the social environment in real life.
The identity field verification adopts a questionnaire form, the identity and occupation of the user in life are determined, and further the favorite and interested fields of the user can be confirmed through the questionnaire form;
the identity field can be verified in a real-name authentication mode, namely, the identity field can be verified by inputting the identity card number and name of the user and face recognition, the identity field can also be authenticated by binding a mobile phone number, and the real-name authentication gateway is connected to a CA mechanism for identity verification and returns a verification result; the real-name authentication gateway is connected to a public security population base for identity verification and returns a verification result; and the real-name authentication gateway receives the verification result and the verification result.
The data length of the identity field is 20 bits, and the length of the USB Key serial number is 16 bits; and constructing a 20-bit first key of the Trivium algorithm by adopting a first formula according to the first 4-bit data in the data of the identity field and the 16-bit serial number, and acquiring the data of the 16-bit output field in the Trivium algorithm by adopting the 20-bit first key.
The first formula is:
the uuidi is the first 4 bits of data of the identity field, the serial number i-4 is the 16-bit serial number of the USB Key, and the Key1 is the first Key of 20 bits. And comparing the data of the 16-bit output field with the data of the 16 bits of the identity field except the first 4 bits, if the data of the 16-bit output field is the same as the data of the 16 bits of the identity field except the first 4 bits, the identity verification is successful, otherwise, the identity verification fails.
According to the 16-bit data of the identity field except the first 4 bits and the first 4-bit data of the first field, constructing a 20-bit second key of the Trivium algorithm by adopting a second formula, and decrypting the data of the first field except the first 4 bits by adopting the 20-bit second key to obtain an actual value of the data of the first field except the first 4 bits;
the second formula is:
wherein uuidi-4 is 16 bits of data except the first 4 bits of the identity field, data i is the first 4 bits of data of the first field, and key2 is the 20 bits of second key.
S2, deep analysis of user questions:
judging whether the user question only needs fuzzy answer or accurate answer;
the fuzzy answer and the accurate answer options are set in the search interface, the fuzzy answer is that keywords are extracted according to the questioning content to search corresponding answers in a questioning and answering database, and the accurate answer is that the relevant fields are limited on the basis of the user role attributes to search corresponding answers in the questioning and answering database.
Obtaining a conversation record, and converting the conversation content into a text form so as to construct a question-answer document; and performing word segmentation, stop word removal and id number allocation operations on the question and answer document to obtain a data set. And taking the data set as the input of the deep learning network model, training the deep learning network model for multiple times, and continuously reducing the entropy values of the true value and the predicted value so as to achieve convergence and form the question-answering system model. The question-answering system trained by adopting the Seq2Seq framework and the Attention mechanism can more accurately understand the semantics of the questions, not only contains the context information of the conversation, but also blends the historical conversation information into the current conversation, thereby being capable of accurately answering the questions.
Taking a translation task as an example, translating "I love machine learning" into "I love machine learning", I "in the first step of decoder output sequence, we want to pay attention to" I "in the input sequence, and translate" I "into" I "; in the third step, we want to focus on "machine" and translate to "machine".
The same word in the source sequence may be output differently in the output sequence, depending on the scene. For example, "I" might be translated into "I" and "me" might be translated. This is a word ambiguity, which is a phenomenon that a word can be mapped into multiple words, and is broadly referred to as "word ambiguity". One effective way to solve the "word ambiguity" problem is to generate word vectors by referring to context information of the source sequence, i.e. context information, and similarly for chinese, generating word vectors by context information.
S3, searching the corresponding feedback answers in the database based on the questions:
when an answer request carrying a target question is received, semantic proximity of each answer to the target question is respectively determined according to the target question, each answer in an answer query library, a question semantic extraction formula, an answer semantic extraction formula, a training value of a question semantic extraction parameter and a training value of an answer semantic extraction parameter. And selecting the target answers from the answers according to the semantic proximity of the answers to the target question and the text proximity of the answers to the target question, and feeding back the answer request.
The database index is the key word of the question proposed by the user, the key word of the question content is used as the index to search the corresponding answer in the question-answer database for feedback, and the database index is created for each structural query key word of the database. Determining field information of each query condition field in a keyword statement; calculating the weight of each query condition field according to the field information of each query condition field; and creating a database index for the keyword according to the weight of each query condition field.
S4, response evaluation:
and constructing an evaluation model, acquiring a plurality of candidate answers from the answer text according to the query text by the evaluation model, and evaluating the candidate answers through a convex neural network. And obtaining the optimal answer in the multiple candidate answers according to the evaluation result.
The user receives the corresponding feedback answers of the questions, the feedback answers are sorted according to the answer priority of the corresponding user role when the feedback answers have a plurality of answers, and then the rest answers are arranged, wherein the rest answers are the database contents of which the answers are not consistent with the attributes of the user role.
S5, deep learning of the database:
performing question-answer database model training based on the result of evaluating the feedback answers in the step S4, and perfecting and adjusting question-answer word materials in the question-answer database through a deep learning technology;
and analyzing the characteristic information of the target initialization neural network in preset neural network big data or a database to obtain an analysis result.
The target initialization network acquisition module is used for determining a target initialization neural network according to the analysis result; and the neural network deep learning module is used for training a target initialization neural network through the training data to obtain a deep learning target neural network.
For example: with the development of the times, new words and sentences often appear, or old words and sentences are added with other meanings in the background of the times, so the words and sentences need to be adjusted and supplemented in a question and answer database.
The model structure for establishing the question-answer database training model is divided into three layers:
the method comprises an input layer, a hidden layer and a softmax layer, wherein onehot vectors are input into the input layer, a weight matrix w is arranged between the input layer and the hidden layer, the hidden layer is set to be 300-dimensional features, the weight matrix w and the onehot vectors are multiplied to obtain a 300-dimensional feature vector, and then probability output is obtained through a wx + b function and then softmax.
The parameters set during model training are as follows: imbedding _ size: 100, skip _ window: 5, num _ skips: 2, num _ steps: 100000, num _ sampled: 64, vocab _ size: 50000, learning _ rate: 0.0001, epoch: 100, batch _ size: 100.
an intelligent question-answering system comprising: at least one memory and at least one processor;
at least one memory for storing a machine readable program;
at least one processor for invoking the machine readable program to perform an intelligent question answering method.
The above embodiments are only specific examples of the present invention, and the scope of the present invention includes but is not limited to the above embodiments, and any suitable changes or substitutions that are consistent with the claims of the intelligent question answering method and system of the present invention and are made by those skilled in the art should fall within the scope of the present invention.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (10)
1. An intelligent question-answering method is characterized by comprising the following steps:
s1, analyzing the role of the user;
s2, deep analysis of user problems;
s3, searching a corresponding feedback answer in the database based on the question;
s4, response evaluation;
and S5, deep learning of the database.
2. The intelligent question answering method according to claim 1, wherein in step S1, the user question is subjected to identity field verification to determine the role attribute of the user, the identity field verification is performed in a questionnaire form or by real-name authentication;
the real-name authentication gateway is connected to a CA mechanism for identity authentication and returns an authentication result; the real-name authentication gateway is connected to a public security population base for identity verification and returns a verification result; and the real-name authentication gateway receives the verification result and the verification result.
3. The intelligent question answering method according to claim 2, wherein the data length of the identity field is 20 bits, the length of the USB Key serial number is 16 bits, a first 20-bit Key of the Trivium algorithm is constructed by using a first formula according to the first 4-bit data in the data of the identity field and the 16-bit serial number, and the data of the 16-bit output field in the Trivium algorithm is obtained by using the 20-bit Key;
the first formula is:
wherein uuidi is the first 4 bits of data of the identity field, serial number i-4 is the 16 bits of serial number of the USB Key, and Key1 is the 20 bits of first Key; and comparing the data of the 16-bit output field with the data of the 16 bits of the identity field except the first 4 bits, if the data of the 16-bit output field is the same as the data of the 16 bits of the identity field except the first 4 bits, the identity verification is successful, otherwise, the identity verification fails.
4. The intelligent question answering method according to claim 3, wherein a second 20-bit key of Trivium algorithm is constructed according to the first 4-bit data of the first field and the 16-bit data of the identity field of the first 4 bits, and the second 20-bit key is used to decrypt the data of the first 4-bit field to obtain the actual value of the data of the first 4-bit field;
the second formula is:
wherein uuidi-4 is 16 bits of data except the first 4 bits of the identity field, data i is the first 4 bits of data of the first field, and key2 is the 20 bits of second key.
5. The intelligent question answering method according to claim 4, wherein in step S2, fuzzy answers and accurate answers are set on a search interface, conversation records are obtained, conversation contents are converted into text forms, question answering documents are constructed, word segmentation, stop word removal and id number assignment are performed on the question answering documents, so that data sets are obtained, the data sets are used as input of a deep learning network model, and the deep learning network model is trained for multiple times to form a question answering system model.
6. The intelligent question answering method according to claim 5, wherein in step S3, when an answer request carrying a target question is received, semantic proximity of each answer to the target question is respectively determined according to the target question, each answer in an answer query library, a question semantic extraction formula, an answer semantic extraction formula, a training value of a question semantic extraction parameter and a training value of an answer semantic extraction parameter; and selecting the target answers from the answers according to the semantic proximity of the answers to the target question and the text proximity of the answers to the target question, and feeding back the answer request.
7. The intelligent question-answering method according to claim 6, wherein in step S4, an evaluation model is constructed, the evaluation model obtains a plurality of candidate answers from an answer text according to a question text, and evaluates the candidate answers through a convex neural network; and obtaining the optimal answer in the multiple candidate answers according to the evaluation result.
8. The intelligent question answering method according to claim 7, wherein in step S5, based on step S4, a question answering database model is trained, and a question answering word stock in the question answering database is perfected and adjusted through a deep learning technology;
and analyzing the characteristic information of the target initialization neural network in preset neural network big data or a database to obtain an analysis result.
9. The intelligent question answering method according to claim 8, wherein the model structure of the question answering database model is divided into three layers:
the method comprises an input layer, a hidden layer and a softmax layer, wherein onehot vectors are input into the input layer, a weight matrix w is arranged between the input layer and the hidden layer, the hidden layer is set to be 300-dimensional features, the weight matrix w and the onehot vectors are multiplied to obtain a 300-dimensional feature vector, and then probability output is obtained through a wx + b function and then softmax.
10. An intelligent question-answering system, comprising: at least one memory and at least one processor;
the at least one memory to store a machine readable program;
the at least one processor, configured to invoke the machine readable program to perform the method of any of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210560246.9A CN114860883A (en) | 2022-05-23 | 2022-05-23 | Intelligent question and answer method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210560246.9A CN114860883A (en) | 2022-05-23 | 2022-05-23 | Intelligent question and answer method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114860883A true CN114860883A (en) | 2022-08-05 |
Family
ID=82639708
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210560246.9A Pending CN114860883A (en) | 2022-05-23 | 2022-05-23 | Intelligent question and answer method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114860883A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI842359B (en) | 2022-12-14 | 2024-05-11 | 大陸商鼎捷軟件股份有限公司 | Question-and-anser system and operating method thereof |
-
2022
- 2022-05-23 CN CN202210560246.9A patent/CN114860883A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI842359B (en) | 2022-12-14 | 2024-05-11 | 大陸商鼎捷軟件股份有限公司 | Question-and-anser system and operating method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110765244B (en) | Method, device, computer equipment and storage medium for obtaining answering operation | |
| CN109241251B (en) | Conversation interaction method | |
| CN108153876B (en) | Intelligent question and answer method and system | |
| CN108052583B (en) | E-commerce ontology construction method | |
| WO2021159632A1 (en) | Intelligent questioning and answering method and apparatus, computer device, and computer storage medium | |
| CN111753060A (en) | Information retrieval method, device, equipment and computer readable storage medium | |
| CN105808590B (en) | Search engine implementation method, searching method and device | |
| CN110765257A (en) | Intelligent consulting system of law of knowledge map driving type | |
| CN110096577A (en) | From the intention of abnormal profile data prediction user | |
| CN112800170A (en) | Question matching method and device and question reply method and device | |
| CN109460459B (en) | Log learning-based dialogue system automatic optimization method | |
| CN104836720A (en) | Method for performing information recommendation in interactive communication, and device | |
| CN111783474A (en) | Comment text viewpoint information processing method and device and storage medium | |
| CN110866089B (en) | Robot knowledge base construction system and method based on synonymous multi-context analysis | |
| CN111309887B (en) | Method and system for training text key content extraction model | |
| CN113094578A (en) | Deep learning-based content recommendation method, device, equipment and storage medium | |
| CN112307164A (en) | Information recommendation method and device, computer equipment and storage medium | |
| CN113569001A (en) | Text processing method and device, computer equipment and computer readable storage medium | |
| CN112287090A (en) | Financial question asking back method and system based on knowledge graph | |
| CN115840812A (en) | Method and system for intelligently matching enterprises according to policy text | |
| CN111104492B (en) | Civil aviation field automatic question and answer method based on layering Attention mechanism | |
| CN113392305A (en) | Keyword extraction method and device, electronic equipment and computer storage medium | |
| TWI734085B (en) | Dialogue system using intention detection ensemble learning and method thereof | |
| CN111460114A (en) | Retrieval method, device, equipment and computer readable storage medium | |
| CN115577080A (en) | Question reply matching method, system, server and storage medium |
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 |