CN117271755B - Custom closed-loop rule engine management control method based on artificial intelligence - Google Patents
Custom closed-loop rule engine management control method based on artificial intelligence Download PDFInfo
- Publication number
- CN117271755B CN117271755B CN202311548638.4A CN202311548638A CN117271755B CN 117271755 B CN117271755 B CN 117271755B CN 202311548638 A CN202311548638 A CN 202311548638A CN 117271755 B CN117271755 B CN 117271755B
- Authority
- CN
- China
- Prior art keywords
- text
- semantic
- candidate document
- expansion
- representing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 33
- 239000013598 vector Substances 0.000 claims abstract description 185
- 238000004364 calculation method Methods 0.000 claims abstract description 33
- 238000012545 processing Methods 0.000 claims abstract description 25
- 238000009223 counseling Methods 0.000 claims description 37
- 238000007781 pre-processing Methods 0.000 claims description 19
- 238000000605 extraction Methods 0.000 claims description 17
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000000546 chi-square test Methods 0.000 claims description 4
- 238000010187 selection method Methods 0.000 claims description 2
- 238000004891 communication Methods 0.000 description 8
- 239000003814 drug Substances 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000001668 ameliorated effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000000007 visual effect Effects 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
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/3332—Query translation
- G06F16/3334—Selection or weighting of terms from queries, including natural language queries
-
- 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
- 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
- G06F16/3347—Query execution using vector based model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention relates to the technical field of intelligent customer complaints, and discloses a customer complaint closed-loop rule engine management control method based on artificial intelligence, which comprises the following steps: semantic coding is carried out on the pre-processed consultation problem text data and candidate document text data; characteristic expansion is carried out on the semantic vectors obtained by encoding by adopting a customer complaint closed-loop rule, and additional semantic features of association rules are established on the semantic vectors after characteristic expansion of candidate document texts; and carrying out similarity calculation on the semantic vectors after the association expansion of the candidate document texts and the semantic vectors after the feature expansion of the consultation question texts, and selecting the candidate document texts with high similarity to carry out customer service answers. The invention carries out the feature expansion processing of the approximate word coding vector expansion and the question-answer association expansion on the semantic vector, expands the semantic range represented by the limited-length question text, improves the similarity calculation accuracy of the consultation question text and the candidate document text, and further realizes the customer service accurate reply.
Description
Technical Field
The invention relates to the field of intelligent customer complaints, in particular to a customer complaint closed-loop rule engine management control method based on artificial intelligence.
Background
An intelligent customer service system is an automated system that utilizes artificial intelligence techniques to help users solve problems and provide information. In modern society, with popularization of the internet and explosive growth of information amount, people have increasingly high demands for obtaining accurate and rapid information. The traditional manual customer service has the problems of low service efficiency, high cost, time and space limitation and the like. Therefore, these problems can be effectively ameliorated by introducing intelligent customer service systems. At present, many research results of intelligent customer service are constructed by a semantic understanding and reasoning model, so that an intelligent customer service system can better understand user questions and give accurate answers. However, the existing research still has problems, because of rich english text data resources, the related research of knowledge base questions and answers is mainly concentrated on english questions and answers, and the knowledge base questions and answers in chinese and other languages have less data, so that intelligent and accurate customer service problem processing in a multi-language environment cannot be realized, and poor user experience is caused. Aiming at the problem, the invention provides a custom closed-loop rule engine management control method based on artificial intelligence, which realizes the self-adaptive customer service accurate reply in different language environments.
Disclosure of Invention
In view of this, the present invention provides a custom closed-loop rule engine management control method based on artificial intelligence, which aims to: 1) Generating candidate document texts of multiple languages for custom processing, realizing multi-language custom service, performing word segmentation processing on the document texts by adopting a mode of calculating the adjacent co-occurrence probability among the characters, generating feature selection weight of each phrase by adopting a mode based on chi-square test, improving the coding weight of key phrases, realizing the coding pretreatment of the document texts, extracting semantic vectors of the document texts by adopting a mode of combining self-attention, and realizing the semantic feature extraction of the document texts; 2) The method comprises the steps of performing feature expansion on a semantic vector obtained by encoding by adopting a custom closed-loop rule, expanding the semantic vector by using the encoding vector of an approximate word in a candidate document, realizing closed-loop expansion of the query question and the encoding vector of the approximate word of the candidate document text, avoiding the situation that an answer to the query question cannot be given through the semantic vector, and simultaneously, using the encoding vector of the approximate word in the candidate document, avoiding the influence of an irrelevant encoding vector on the retrieval precision, establishing additional semantic features of the association rule on the semantic vector after the feature expansion of the candidate document text by adopting a mode of combining the association rule of the question and the answer in the candidate document text, obtaining the semantic vector after the association expansion of the candidate document text, performing similarity calculation on the semantic vector after the association expansion of the candidate document text and the semantic vector after the feature expansion of the query question text, selecting a corresponding candidate document text according to a similarity calculation result, and performing customer service answer extraction, obtaining the semantic vector combined with the encoding vector expansion of the approximate word and the question-association expansion, expanding the semantic range represented by the limited-question text, and improving the accuracy of the similarity calculation of the candidate document text, thereby realizing customer service precision.
In order to achieve the above purpose, the custom closed-loop rule engine management control method based on artificial intelligence provided by the invention comprises the following steps:
s1: generating candidate document texts in multiple languages for customer complaint processing to form a database, acquiring consultation problem texts, preprocessing the consultation problem texts and the candidate document texts stored in the database to obtain preprocessed consultation problem text data and candidate document text data, wherein a characteristic selection method based on chi-square test is a specific implementation method of preprocessing;
s2: semantic coding is carried out on the pre-processed consultation problem text data and candidate document text data, so that semantic vectors of consultation problems and candidate document texts are obtained respectively;
s3: characteristic expansion is carried out on the semantic vectors obtained by encoding by adopting a customer complaint closed-loop rule, and additional semantic features of association rules are established on the semantic vectors obtained by characteristic expansion of the candidate document text, so that the semantic vectors obtained by association expansion of the candidate document text are obtained;
s4: and carrying out similarity calculation on the semantic vectors after the association expansion of the candidate document texts and the semantic vectors after the feature expansion of the consultation question texts, and selecting the corresponding candidate document texts according to the similarity calculation result to carry out customer service answer extraction.
As a further improvement of the present invention:
optionally, generating candidate document texts of multiple languages for customer complaint processing in the step S1 to form a database, and acquiring consultation problem texts, including:
generating customer service answer templates for customer complaint processing and question templates corresponding to the customer service answer templates, converting the generated customer service answer templates and the question templates corresponding to the customer service answer templates into multiple languages, forming candidate document texts of the multiple languages for customer complaint processing, and forming a database by the candidate document texts, wherein the candidate document texts in the database are collected as follows:
wherein:
representing candidate document text sequences obtained by performing multiple language conversion on an nth customer service answer template and a question template, wherein N represents the total number of generated customer service answer templates;
representing the generated nth customer service answer template and candidate document text corresponding to the question template,/->，Representing the generated nth customer service answer template, < ->Template for representing customer service answer>A corresponding problem template;
representing the customer service answer template->Question template->Converting into candidate document text obtained in the mth language, M representing the number of language types,/and- >，/>Representing the customer service answer template->To a conversion result of the mth language,representing a question template->Converting the language into a conversion result of the m-th language;
acquiring consultation problem text and identifying language types of the consultation problem text, wherein the expression form of the consultation problem text isS represents the language category of the consultation question text.
Optionally, in the step S1, the consultation problem text and the candidate document text stored in the database are preprocessed, so as to obtain preprocessed consultation problem text data and candidate document text data, which includes:
obtaining candidate document text sequences of the s-th language from a database:
wherein:
representing an nth candidate document text obtained by converting an nth customer service answer template and a question template into an nth language;
preprocessing consultation problem text and acquired candidate document text, wherein the consultation problem textThe pretreatment flow of (2) is as follows:
s11: calculating advisory problem textThe adjacent co-occurrence probability between different words in (a) is as follows:
wherein:
representing the probability of co-occurrence of the proximity between text a and text b;
the number of times that the word a and the word b appear in the text to be preprocessed is respectively represented;
Representing the probability that the left adjacent word is a in the case that word b appears;
if it isAbove the preset threshold, the characters a and a are combinedThe text b is used as a group of phrases, and the consultation question text is +.>A representation divided into a plurality of phrases; in the embodiment of the invention, the words comprise independent words and phrases formed by the words;
s12: calculating advisory problem textFeature selection weights of different phrases in the database, wherein the question text is consulted +.>Is->The feature selection weight calculation formula of (1) is as follows:
wherein:
text representing counseling questions +.>Is->Feature selection weights of ∈10->Text representing counseling questions +.>Middle->No repeated phrase->Text representing counseling questions +.>The total number of the phrase is not repeated;
express phrase->Text +.>Is a frequency of occurrence in the first and second embodiments;
text representing counseling questions +.>The number of sentences in (a);
text representing counseling questions +.>The Chinese medicine contains phrase->Is>Text representing counseling questions +.>The Chinese medicine contains no phrase->Is the number of sentences;
s13: encoding and representing different phrases by adopting a single-heat encoding mode, taking the feature selection weight of the phrases as the weight of the encoding and representing, and generating consultation problem textsIs encoded representation vector of (1) as advisory problem text +. >Is a result of the pretreatment of:
wherein:
text representing counseling questions +.>I.e. pre-processed advisory problem text data;
text representing counseling questions +.>The weighted code of the j-th phrase in the list represents the result, L represents the consultation question text +.>Is a word group. In the embodiment of the invention, the obtained candidate document text is preprocessed in the same preprocessing mode as the consultation question text, wherein the customer service answer templates and the question templates in the candidate document text are preprocessed respectively, and the preprocessing process is not repeated here.
Optionally, in the step S2, the semantic encoding is performed on the pre-processed consultation problem text data and the candidate document text data, including:
semantic coding is carried out on the preprocessed consultation problem text data and the candidate document text data, wherein the semantic coding flow of the preprocessed consultation problem text data is as follows:
s21: generating advisory problem textAttention weight of each phrase:
wherein:
text representing counseling questions +.>Attention weight of the j-th phrase in (a);
respectively a parameter matrix;
t represents a transpose;
representing the length of the weighted code representation result of the phrase;
S22: pre-processed consultation problem text data based on attention weight of phraseAnd (3) weighting:
wherein:
text data representing pre-processed counseling questions +.>Is a result of the weighting process;
s23: generating advisory problem textSemantic vector +.>：
Wherein:
representing a parameter matrix for generating a question text semantic vector; in the embodiment of the present invention, < > a->Semantic vector for generating consultation question text and question template,/-question>And the semantic vector is used for generating customer service answer templates.
In the embodiment of the invention, the obtained candidate document text is semantically encoded by adopting the semantic encoding mode which is the same as that of the consultation question text, so that the customer service answer template and the semantic vector of the question template in the candidate document text are respectively obtained.
Optionally, in the step S3, feature expansion is performed on the semantic vector obtained by encoding by adopting a custom closed-loop rule, including:
characteristic expansion is carried out on the semantic vector obtained by encoding by adopting a custom closed-loop rule, wherein the text of the problem is consultedSemantic vector +.>The characteristic expansion flow of (1) is as follows:
s31: calculating semantic vectorsCosine similarity of each semantic feature and the result of the single-hot coding representation of different phrases, wherein the semantic vector +. >Is expressed as：
Wherein:
representing semantic vector +.>The j-th semantic feature of (i.e. consultation question text +.>Semantic features of the j-th phrase;
s32: selecting the single-hot coding representation result of k phrases with highest cosine similarity to expand semantic features, wherein the semantic featuresThe expansion result of (2) is:
wherein:
representation and semantic feature->The single thermal code of k phrases with highest cosine similarity represents the result;
representing semantic features->Is a result of the expansion of (a);
s33: semantic vectors after feature expansion are formed: wherein:
text representing counseling questions +.>Semantic vectors after feature expansion;
carrying out feature expansion on the semantic vectors of the obtained candidate document text by adopting a feature expansion mode which is the same as that of the consultation question text, and respectively obtaining customer service answer templates in the candidate document text and semantic vectors after feature expansion of the question templates: wherein:
representing the semantic vector of the nth candidate document text after feature expansion;
semantic vector representing feature expansion of nth customer service answer template,/item>And representing the semantic vector after feature expansion of the question template corresponding to the nth customer service answer template.
Optionally, in the step S3, additional semantic features of an association rule are established for the semantic vectors after feature expansion of the candidate document text, so as to obtain associated semantic vectors after the association expansion of the candidate document text, including:
Establishing additional semantic features of association rules for semantic vectors of problem templates in candidate document texts after feature expansion to obtain semantic vectors of problem templates in candidate document texts after association expansion, wherein the additional semantic features are specialPost sign expansion semantic vectorThe associated expansion formula of (2) is:
wherein:
representing post feature expansion semantic vector->The related expansion result of the question template corresponding to the nth answer template is the related expanded semantic vector;
representing a convolution processing operation;
an exponential function based on a natural constant is represented.
Optionally, in the step S4, similarity calculation is performed on the semantic vector after the association expansion of the candidate document text and the semantic vector after the feature expansion of the consultation problem text, including:
similarity calculation is carried out on the semantic vector after the association expansion of the candidate document text and the semantic vector after the feature expansion of the consultation problem text, wherein the semantic vector after the association expansionSemantic vector after feature expansion->The similarity calculation formula of (2) is:
wherein:
representing post-associative expansion semantic vector->Semantic vector after feature expansion->Similarity of (2);
representing the reordering of the vectors of each column in the semantic vector according to the descending order of the attention weights of the phrases; in the embodiment of the invention, the j-th column of the semantic vector represents the semantic information of the j-th phrase in the text document;
Representing the L1 norm.
Optionally, in the step S4, selecting a corresponding candidate document text according to the similarity calculation result to perform customer service answer extraction, including:
selecting a customer service answer template in a candidate document text corresponding to the semantic vector after the association expansion with the highest similarity as a customer service answer extraction result, and taking the customer service answer extraction result as a consultation question textIs a reply to the user.
The invention provides a custom closed-loop rule engine management control method based on artificial intelligence, which comprises the following steps:
the text preprocessing module is used for generating candidate document texts in multiple languages for customer complaint processing to form a database, acquiring consultation problem texts, and preprocessing the consultation problem texts and the candidate document texts stored in the database;
the semantic coding expansion module is used for carrying out semantic coding on the preprocessed consultation problem text data and the candidate document text data to respectively obtain semantic vectors of the consultation problem and the candidate document text, carrying out feature expansion on the semantic vectors obtained by coding by adopting a custom closed-loop rule, and establishing additional semantic features of association rules on the semantic vectors after feature expansion of the candidate document text to obtain associated expanded semantic vectors of the candidate document text;
And the customer complaint replying device is used for carrying out similarity calculation on the semantic vectors after the association expansion of the candidate document text and the semantic vectors after the feature expansion of the consultation question text, and selecting the corresponding candidate document text according to the similarity calculation result to carry out customer service answer extraction.
In order to solve the above-mentioned problems, the present invention also provides an electronic apparatus including:
a memory storing at least one instruction;
the communication interface is used for realizing the communication of the electronic equipment; and
And the processor executes the instructions stored in the memory to realize the custom closed-loop rule engine management control method based on the artificial intelligence.
In order to solve the above problems, the present invention further provides a computer readable storage medium, in which at least one instruction is stored, the at least one instruction being executed by a processor in an electronic device to implement the above-mentioned artificial intelligence-based complaint closed-loop rule engine management control method.
Compared with the prior art, the invention provides a custom closed-loop rule engine management control method based on artificial intelligence, and the technology has the following advantages:
firstly, the proposal provides a text coding mode for preprocessing consultation problem text and acquired candidate document text, wherein the consultation problem text The pretreatment flow of (2) is as follows: calculating consultation problem text->Adjacent co-occurrence probabilities between different words in a document, whereinThe probability of co-occurrence of the adjacency between the word a and the word b is:
wherein: />Representing the probability of co-occurrence of the proximity between text a and text b; />The number of times that the word a and the word b appear in the text to be preprocessed is respectively represented; />Representing the probability that the left adjacent word is a in the case that word b appears; if->If the text is higher than the preset threshold value, the text a and the text b are used as a group of phrases, and the consultation problem text is +.>A representation divided into a plurality of phrases; in the embodiment of the invention, the words comprise independent words and phrases formed by the words; calculating consultation problem text->Feature selection weights of different phrases in the database, wherein the question text is consulted +.>Is->The feature selection weight calculation formula of (1) is as follows:
wherein: />Representing the consultationQuestion text of inquiry->Is->Feature selection weights of ∈10->Text representing counseling questions +.>Middle->No repeated phrase->Text representing counseling questions +.>The total number of the phrase is not repeated; />Express phrase->Text +.>Is a frequency of occurrence in the first and second embodiments; />Text representing counseling questions +.>The number of sentences in (a); / >Text representing counseling questions +.>The Chinese medicine contains phrase->Is>Text representing counseling questions +.>The Chinese medicine contains no phrase->Is the number of sentences; coding and representing different phrases by adopting a single-hot coding mode, taking the feature selection weight of the phrases as the weight of the coding and representing, and generating consultation question text +.>Coded representation vectors as advisory problem textIs a result of the pretreatment of:
wherein: />Text representing counseling questions +.>I.e. pre-processed advisory problem text data; />Text representing counseling questions +.>The weighted code of the j-th phrase in the list represents the result, L represents the consultation question text +.>Is a word group. The scheme generates candidate document text of multiple languages for complaint processing to realize multi-language complaint service, and calculates the neighborhood between charactersThe method comprises the steps of performing word segmentation on a document text in a near co-occurrence probability mode, generating feature selection weight of each phrase in a chi-square test mode, improving coding weight of key phrases, realizing coding pretreatment of the document text, extracting semantic vectors of the document text in a self-attention combined mode, and realizing semantic feature extraction of the document text.
Meanwhile, the scheme provides a semantic vector expansion mode, and characteristic expansion is carried out on the semantic vector obtained by encoding by adopting a custom closed-loop rule, wherein the text of the problem is consulted Semantic vector +.>The characteristic expansion flow of (1) is as follows: calculating semantic vector +.>Cosine similarity of each semantic feature and the result of the single-hot coding representation of different phrases, wherein the semantic vector +.>Is expressed as:
wherein: />Representing semantic vector +.>The j-th semantic feature of (i.e. consultation question text +.>Semantic features of the j-th phrase; selecting the single-hot coding representation result of k phrases with highest cosine similarity to expand semantic features, wherein the semantic features are +.>The expansion result of (2) is:
wherein: />Representation and semantic feature->The single thermal code of k phrases with highest cosine similarity represents the result; />Representing semantic features->Is a result of the expansion of (a); semantic vectors after feature expansion are formed:
wherein: />Text representing counseling questions +.>Semantic vectors after feature expansion; carrying out feature expansion on the semantic vectors of the obtained candidate document text by adopting a feature expansion mode which is the same as that of the consultation question text, and respectively obtaining customer service answer templates in the candidate document text and semantic vectors after feature expansion of the question templates: /> Wherein: />Representing the semantic vector of the nth candidate document text after feature expansion; />Semantic vector representing feature expansion of nth customer service answer template,/item >And representing the semantic vector after feature expansion of the question template corresponding to the nth customer service answer template. Establishing additional semantic features of association rules for semantic vectors of the problem templates in the candidate document text after feature expansion to obtain the semantic vectors of the problem templates in the candidate document text after association expansion, wherein the semantic vectors of the problem templates in the candidate document text after feature expansion are +.>The associated expansion formula of (2) is:
wherein: />Representing post feature expansion semantic vector->The related expansion result of the question template corresponding to the nth answer template is the related expanded semantic vector; />Representing a convolution processing operation; />An exponential function based on a natural constant is represented. The scheme adopts a custom closed-loop rule to perform feature expansion on semantic vectors obtained by encoding, uses the encoding vectors of approximate words in candidate documents to expand the semantic vectors, realizes closed-loop expansion on the encoding vectors of the approximate words of consultation questions and candidate document texts, avoids the situation that accurate answers cannot be given because consultation questions cannot be retrieved through the semantic vectors, simultaneously uses the encoding vectors of the approximate words in the candidate documents, avoids expansion of irrelevant encoding vectors from influencing retrieval precision, and combines the questions in the candidate document texts with the retrieval precision And establishing additional semantic features of the association rule for the semantic vectors after feature expansion of the candidate document text in the manner of the association rule of the answer, obtaining the semantic vectors after the association expansion of the candidate document text, carrying out similarity calculation on the semantic vectors after the association expansion of the candidate document text and the semantic vectors after the feature expansion of the consultation question text, selecting the corresponding candidate document text according to a similarity calculation result, and carrying out customer service answer extraction to obtain the semantic vectors combined with the similar word code vector expansion and the question-answer association expansion, thereby expanding the semantic range represented by the limited-length question text, improving the similarity calculation accuracy of the consultation question text and the candidate document text, and further realizing customer service accurate reply.
Drawings
FIG. 1 is a flow chart of a custom closed-loop rule engine management control method based on artificial intelligence according to an embodiment of the invention;
FIG. 2 is a functional block diagram of a custom closed-loop rule engine management control method based on artificial intelligence according to an embodiment of the present invention;
wherein: 100. a custom closed-loop rule engine management control system based on artificial intelligence; 101. a text preprocessing module; 102. a semantic code expansion module; 103. a customer complaint replying device;
FIG. 3 is a schematic structural diagram of an electronic device implementing a custom closed-loop rule engine management control method based on artificial intelligence according to an embodiment of the present invention;
wherein: 1. an electronic device; 10. a processor; 11. a memory; 12. a program; 13. a communication interface;
the achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The embodiment of the application provides a custom closed-loop rule engine management control method based on artificial intelligence. The execution body of the artificial intelligence-based complaint closed-loop rule engine management control method comprises, but is not limited to, at least one of a server, a terminal and the like which can be configured to execute the method provided by the embodiment of the application. In other words, the artificial intelligence based complaint closed-loop rule engine management control method may be performed by software or hardware installed in a terminal device or a server device, and the software may be a blockchain platform. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like.
Example 1:
the custom closed-loop rule engine management control method based on artificial intelligence comprises the following steps:
s1: generating candidate document texts in multiple languages for customer complaint processing to form a database, acquiring consultation problem texts, and preprocessing the consultation problem texts and the candidate document texts stored in the database to obtain preprocessed consultation problem text data and candidate document text data.
The step S1 is to generate candidate document texts of multiple languages for customer complaint processing to form a database, and acquire consultation problem texts, and the method comprises the following steps:
generating customer service answer templates for customer complaint processing and question templates corresponding to the customer service answer templates, converting the generated customer service answer templates and the question templates corresponding to the customer service answer templates into multiple languages, forming candidate document texts of the multiple languages for customer complaint processing, and forming a database by the candidate document texts, wherein the candidate document texts in the database are collected as follows:
wherein:
representing candidate documents obtained by performing multiple language conversion on nth customer service answer templates and question templatesA text sequence, N representing the total number of customer service answer templates generated;
representing the generated nth customer service answer template and candidate document text corresponding to the question template,/- >，Representing the generated nth customer service answer template, < ->Template for representing customer service answer>A corresponding problem template;
representing the customer service answer template->Question template->Converting into candidate document text obtained in the mth language, M representing the number of language types,/and->，/>Representing the customer service answer template->To a conversion result of the mth language,representing a question template->Converting the language into a conversion result of the m-th language;
acquiring consultation problem text and identifying language types of the consultation problem text, wherein the expression form of the consultation problem text isS represents the language category of the consultation question text.
In the step S1, preprocessing the consultation problem text and the candidate document text stored in the database to obtain preprocessed consultation problem text data and candidate document text data, wherein the method comprises the following steps:
obtaining candidate document text sequences of the s-th language from a database:
wherein:
representing an nth candidate document text obtained by converting an nth customer service answer template and a question template into an nth language;
preprocessing consultation problem text and acquired candidate document text, wherein the consultation problem textThe pretreatment flow of (2) is as follows:
s11: calculating advisory problem text The adjacent co-occurrence probability between different words in (a) is as follows:
wherein:
representing the probability of co-occurrence of the proximity between text a and text b;
the number of times that the word a and the word b appear in the text to be preprocessed is respectively represented;
representing the probability that the left adjacent word is a in the case that word b appears;
if it isIf the text is higher than the preset threshold value, the text a and the text b are used as a group of phrases, and the consultation problem text is +.>A representation divided into a plurality of phrases; in the embodiment of the invention, the words comprise independent words and phrases formed by the words;
s12: calculating advisory problem textFeature selection weights of different phrases in the database, wherein the question text is consulted +.>Is->The feature selection weight calculation formula of (1) is as follows:
wherein:
text representing counseling questions +.>Is->Feature selection weights of ∈10->Text representing counseling questions +.>Middle->No repeated phrase->Text representing counseling questions +.>The total number of the phrase is not repeated;
express phrase->Text +.>Is a frequency of occurrence in the first and second embodiments;
text representing counseling questions +.>The number of sentences in (a);
text representing counseling questions +.>The Chinese medicine contains phrase- >Is>Text representing counseling questions +.>The Chinese medicine contains no phrase->Is the number of sentences;
s13: encoding and representing different phrases by adopting a single-heat encoding mode, taking the feature selection weight of the phrases as the weight of the encoding and representing, and generating consultation problem textsIs encoded representation vector of (1) as advisory problem text +.>Is a result of the pretreatment of:
wherein:
text representing counseling questions +.>I.e. pre-processed advisory problem text data;
text representing counseling questions +.>The weighted code of the j-th phrase in the list represents the result, L represents the consultation question text +.>Is a word group.
S2: and carrying out semantic coding on the preprocessed consultation problem text data and the candidate document text data to respectively obtain semantic vectors of the consultation problem and the candidate document text.
And in the step S2, the preprocessed consultation problem text data and the candidate document text data are subjected to semantic coding, and the method comprises the following steps:
semantic coding is carried out on the preprocessed consultation problem text data and the candidate document text data, wherein the semantic coding flow of the preprocessed consultation problem text data is as follows:
s21: generating advisory problem textAttention weight of each phrase:
Wherein:
text representing counseling questions +.>Attention weight of the j-th phrase in (a);
respectively a parameter matrix;
t represents a transpose;
representing the length of the weighted code representation result of the phrase;
s22: pre-processed consultation problem text data based on attention weight of phraseAnd (3) weighting:
wherein:
text data representing pre-processed counseling questions +.>Is a result of the weighting process;
s23: generating advisory problem textSemantic vector +.>：
Wherein:
representing a parameter matrix used to generate the semantic vector of the question text.
S3: and carrying out feature expansion on the semantic vector obtained by encoding by adopting a custom closed-loop rule, and establishing additional semantic features of association rules on the semantic vector obtained by feature expansion of the candidate document text to obtain the semantic vector obtained by association expansion of the candidate document text.
And S3, performing feature expansion on the semantic vector obtained by encoding by adopting a customer complaint closed-loop rule, wherein the method comprises the following steps:
characteristic expansion is carried out on the semantic vector obtained by encoding by adopting a custom closed-loop rule, wherein the text of the problem is consultedSemantic vector +.>The characteristic expansion flow of (1) is as follows:
s31: calculating semantic vectorsCosine similarity of each semantic feature and the result of the single-hot coding representation of different phrases, wherein the semantic vector +. >Is expressed as:
wherein:
representing semantic vector +.>The j-th semantic feature of (i.e. consultation question text +.>Semantic features of the j-th phrase;
s32: selecting the single-hot coding representation result of k phrases with highest cosine similarity to expand semantic features, wherein the semantic featuresThe expansion result of (2) is:
wherein:
representation and semantic feature->The single thermal code of k phrases with highest cosine similarity represents the result;
representing semantic features->Is a result of the expansion of (a);
s33: semantic vectors after feature expansion are formed:
wherein:
text representing counseling questions +.>Semantic vectors after feature expansion;
carrying out feature expansion on the semantic vectors of the obtained candidate document text by adopting a feature expansion mode which is the same as that of the consultation question text, and respectively obtaining customer service answer templates in the candidate document text and semantic vectors after feature expansion of the question templates: wherein:
representing the semantic vector of the nth candidate document text after feature expansion;
semantic vector representing feature expansion of nth customer service answer template,/item>And representing the semantic vector after feature expansion of the question template corresponding to the nth customer service answer template.
In the step S3, additional semantic features of association rules are established for the semantic vectors after feature expansion of the candidate document text, so as to obtain the semantic vectors after association expansion of the candidate document text, and the method comprises the following steps:
Establishing additional semantic features of association rules for semantic vectors of the problem templates in the candidate document text after feature expansion to obtain associated expanded semantic vectors of the problem templates in the candidate document text, wherein the semantic vectors after feature expansionThe associated expansion formula of (2) is:
wherein:
representing post feature expansion semantic vector->The related expansion result of the question template corresponding to the nth answer template is the related expanded semantic vector;
representing a convolution processing operation;
an exponential function based on a natural constant is represented.
S4: and carrying out similarity calculation on the semantic vectors after the association expansion of the candidate document texts and the semantic vectors after the feature expansion of the consultation question texts, and selecting the corresponding candidate document texts according to the similarity calculation result to carry out customer service answer extraction.
In the step S4, similarity calculation is performed on the semantic vector after the association expansion of the candidate document text and the semantic vector after the feature expansion of the consultation question text, and the method comprises the following steps:
similarity calculation is carried out on the semantic vector after the association expansion of the candidate document text and the semantic vector after the feature expansion of the consultation problem text, wherein the semantic vector after the association expansionSemantic vector after feature expansion- >The similarity calculation formula of (2) is:
wherein:
representing post-associative expansion semantic vector->Semantic vector after feature expansion->Similarity of (2);
representing the reordering of the vectors of each column in the semantic vector according to the descending order of the attention weights of the phrases; in the embodiment of the invention, the j-th column of the semantic vector represents the semantic information of the j-th phrase in the text document;
representing the L1 norm.
And S4, selecting a corresponding candidate document text according to the similarity calculation result to extract customer service answers, wherein the step comprises the following steps:
selecting a customer service answer template in a candidate document text corresponding to the semantic vector after the association expansion with the highest similarity as a customer service answer extraction result, and taking the customer service answer extraction result as a consultation question textIs a reply to the user.
Example 2:
as shown in FIG. 2, a functional block diagram of an artificial intelligence based custom closed-loop rule engine management control method according to an embodiment of the present invention may implement the artificial intelligence based custom closed-loop rule engine management control method according to embodiment 1.
The custom closed-loop rule engine management control system 100 based on artificial intelligence can be installed in electronic equipment. Depending on the functions implemented, the artificial intelligence based custom closed-loop rule engine management control system may include a text preprocessing module 101, a semantic code expansion module 102, and a custom response device 103. The module of the invention, which may also be referred to as a unit, refers to a series of computer program segments, which are stored in the memory of the electronic device, capable of being executed by the processor of the electronic device and of performing a fixed function.
A text preprocessing module 101, configured to generate candidate document texts in multiple languages for customer complaint processing to form a database, obtain a consultation problem text, and perform preprocessing on the consultation problem text and the candidate document text stored in the database;
the semantic code expansion module 102 is configured to perform semantic coding on the preprocessed consultation problem text data and the candidate document text data to respectively obtain semantic vectors of the consultation problem and the candidate document text, perform feature expansion on the semantic vectors obtained by the coding by adopting a custom closed-loop rule, and establish additional semantic features of association rules on the semantic vectors after feature expansion of the candidate document text to obtain associated expanded semantic vectors of the candidate document text;
and the customer complaint replying device 103 is used for carrying out similarity calculation on the semantic vectors after the association expansion of the candidate document text and the semantic vectors after the feature expansion of the consultation question text, and selecting the corresponding candidate document text according to the similarity calculation result to carry out customer service answer extraction.
In detail, the modules in the artificial intelligence-based custom closed-loop rule engine management control system 100 in the embodiment of the present invention use the same technical means as the above-mentioned artificial intelligence-based custom closed-loop rule engine management control method in fig. 1, and can produce the same technical effects, which are not described herein.
Example 3:
fig. 3 is a schematic structural diagram of an electronic device implementing a custom closed-loop rule engine management control method based on artificial intelligence according to an embodiment of the present invention.
The electronic device 1 may comprise a processor 10, a memory 11, a communication interface 13 and a bus, and may further comprise a computer program, such as program 12, stored in the memory 11 and executable on the processor 10.
The memory 11 includes at least one type of readable storage medium, including flash memory, a mobile hard disk, a multimedia card, a card memory (e.g., SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. The memory 11 may in some embodiments be an internal storage unit of the electronic device 1, such as a removable hard disk of the electronic device 1. The memory 11 may in other embodiments also be an external storage device of the electronic device 1, such as a plug-in mobile hard disk, a smart memory card (SmartMediaCard, SMC), a secure digital (SecureDigital, SD) card, a flash card (FlashCard) or the like, provided on the electronic device 1. Further, the memory 11 may also include both an internal storage unit and an external storage device of the electronic device 1. The memory 11 may be used not only for storing application software installed in the electronic device 1 and various types of data, such as codes of the program 12, but also for temporarily storing data that has been output or is to be output.
The processor 10 may be comprised of integrated circuits in some embodiments, for example, a single packaged integrated circuit, or may be comprised of multiple integrated circuits packaged with the same or different functions, including one or more central processing units (CentralProcessingunit, CPU), microprocessors, digital processing chips, graphics processors, a combination of various control chips, and the like. The processor 10 is a control unit (control unit) of the electronic device, connects respective components of the entire electronic device using various interfaces and lines, executes or executes programs or modules (a program 12 for implementing a closed-loop rule, etc.) stored in the memory 11, and invokes data stored in the memory 11 to perform various functions of the electronic device 1 and process the data.
The communication interface 13 may comprise a wired interface and/or a wireless interface (e.g. WI-FI interface, bluetooth interface, etc.), typically used to establish a communication connection between the electronic device 1 and other electronic devices and to enable connection communication between internal components of the electronic device.
The bus may be an Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be classified as an address bus, a data bus, a control bus, etc. The bus is arranged to enable a connection communication between the memory 11 and at least one processor 10 etc.
Fig. 3 shows only an electronic device with components, it being understood by a person skilled in the art that the structure shown in fig. 3 does not constitute a limitation of the electronic device 1, and may comprise fewer or more components than shown, or may combine certain components, or may be arranged in different components.
For example, although not shown, the electronic device 1 may further include a power source (such as a battery) for supplying power to each component, and preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that functions of charge management, discharge management, power consumption management, and the like are implemented through the power management device. The power supply may also include one or more of any of a direct current or alternating current power supply, recharging device, power failure detection circuit, power converter or inverter, power status indicator, etc. The electronic device 1 may further include various sensors, bluetooth modules, wi-Fi modules, etc., which will not be described herein.
The electronic device 1 may optionally further comprise a user interface, which may be a Display, an input unit, such as a Keyboard (Keyboard), or a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (organic light-emitting diode) touch, or the like. The display may also be referred to as a display screen or display unit, as appropriate, for displaying information processed in the electronic device 1 and for displaying a visual user interface.
It should be understood that the embodiments described are for illustrative purposes only and are not limited to this configuration in the scope of the patent application.
It should be noted that, the foregoing reference numerals of the embodiments of the present invention are merely for describing the embodiments, and do not represent the advantages and disadvantages of the embodiments. And the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, apparatus, article or method that comprises the element.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) as described above, comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (6)
1. The custom closed-loop rule engine management control method based on artificial intelligence is characterized by comprising the following steps:
s1: generating candidate document texts in multiple languages for customer complaint processing to form a database, acquiring consultation problem texts, preprocessing the consultation problem texts and the candidate document texts stored in the database to obtain preprocessed consultation problem text data and candidate document text data, wherein a characteristic selection method based on chi-square test is a specific implementation method of preprocessing; s2: semantic coding is carried out on the pre-processed consultation problem text data and candidate document text data, so that semantic vectors of consultation problems and candidate document texts are obtained respectively;
s3: characteristic expansion is carried out on the semantic vectors obtained by encoding by adopting a customer complaint closed-loop rule, and additional semantic features of association rules are established on the semantic vectors obtained by characteristic expansion of the candidate document text, so that the semantic vectors obtained by association expansion of the candidate document text are obtained;
S4: carrying out similarity calculation on the semantic vectors after the association expansion of the candidate document texts and the semantic vectors after the feature expansion of the consultation question texts, and selecting the corresponding candidate document texts according to the similarity calculation result to carry out customer service answer extraction;
and S3, performing feature expansion on the semantic vector obtained by encoding by adopting a customer complaint closed-loop rule, wherein the method comprises the following steps:
characteristic expansion is carried out on the semantic vector obtained by encoding by adopting a custom closed-loop rule, wherein the text of the problem is consultedSemantic vector +.>The characteristic expansion flow of (1) is as follows:
s31: calculating semantic vectorsCosine similarity of each semantic feature and the result of the single-hot coding representation of different phrases, wherein the semantic vector +.>Is expressed as:
wherein:
representing semantic vector +.>The j-th semantic feature of (i.e. consultation question text +.>Semantic features of the j-th phrase +.>Representing the number of semantic features;
s32: selecting the single-hot coding representation result of k phrases with highest cosine similarity to expand semantic features, wherein the semantic featuresThe expansion result of (2) is:
wherein:
representation and semantic feature->The single thermal code of k phrases with highest cosine similarity represents the result;
representing semantic features- >Is a result of the expansion of (a);
s33: semantic vectors after feature expansion are formed:
wherein:
text representing counseling questions +.>Semantic vectors after feature expansion;
carrying out feature expansion on the semantic vectors of the obtained candidate document text by adopting a feature expansion mode which is the same as that of the consultation question text, and respectively obtaining customer service answer templates in the candidate document text and semantic vectors after feature expansion of the question templates:
wherein:
features representing the nth candidate document textThe expanded semantic vector;
representing the semantic vector of the N candidate document text after feature expansion;
representing the semantic vector of the nth candidate document text after feature expansion;
semantic vector representing feature expansion of nth customer service answer template,/item>Representing semantic vectors after feature expansion of a question template corresponding to the nth customer service answer template;
in the step S3, additional semantic features of association rules are established for the semantic vectors after feature expansion of the candidate document text, so as to obtain the semantic vectors after association expansion of the candidate document text, and the method comprises the following steps:
establishing additional semantic features of association rules for semantic vectors of the problem templates in the candidate document text after feature expansion to obtain associated expanded semantic vectors of the problem templates in the candidate document text, wherein the semantic vectors after feature expansion The associated expansion formula of (2) is:
wherein:
representing post feature expansion semantic vector->The related expansion result of the question template corresponding to the nth answer template is the related expanded semantic vector;
representing a convolution processing operation;
an exponential function based on a natural constant is represented.
2. The artificial intelligence based complaint closed-loop rule engine management control method as claimed in claim 1, wherein the step S1 of generating a candidate document text composition database of a plurality of languages for complaint processing and obtaining a consultation problem text includes:
generating customer service answer templates for customer complaint processing and question templates corresponding to the customer service answer templates, converting the generated customer service answer templates and the question templates corresponding to the customer service answer templates into multiple languages, forming candidate document texts of the multiple languages for customer complaint processing, and forming a database by the candidate document texts, wherein the candidate document texts in the database are collected as follows:
wherein:
representing candidate document text sequences obtained by performing multiple language conversion on an nth customer service answer template and a question template, wherein N represents the total number of generated customer service answer templates;
representing the generated nth customer service answer template and candidate document text corresponding to the question template,/- >，Representing the generated nth customer service answer template, < ->Template for representing customer service answer>A corresponding problem template;
representing the customer service answer template->Question template->Converting into candidate document text obtained in the mth language, M representing the number of language types,/and->，/>Representing the customer service answer template->Conversion result to the mth language, < >>Representing a question template->Converting the language into a conversion result of the m-th language;
acquiring consultation problem text and identifying language types of the consultation problem text, wherein the expression form of the consultation problem text isS represents the language category of the consultation question text.
3. The artificial intelligence based custom closed-loop rule engine management control method according to claim 2, wherein the step S1 of preprocessing the consultation problem text and the candidate document text stored in the database to obtain preprocessed consultation problem text data and candidate document text data includes:
acquiring candidate document text sequences of the s-th language from a database, and preprocessing consultation problem text and acquired candidate document text, wherein the consultation problem textThe pretreatment flow of (2) is as follows:
s11: calculating advisory problem text The adjacent co-occurrence probability among different characters is used as a phrase;
s12: calculating advisory problem textThe feature selection weights of different phrases in the database;
s13: counseling problem text by combining single-hot coding mode and feature selection weightThe different phrases in the Chinese character string are weighted and encodedCode representation and splicing the weighted coded representation results as consultation question text +.>Is a result of the pretreatment of (a).
4. The artificial intelligence based custom closed-loop rule engine management control method according to claim 3, wherein the step S2 of semantically encoding the pre-processed consultation problem text data and candidate document text data comprises:
semantic coding is carried out on the preprocessed consultation problem text data and the candidate document text data, wherein the semantic coding flow of the preprocessed consultation problem text data is as follows:
s21: generating advisory problem textAttention weight of each phrase:
wherein:text representing counseling questions +.>The code vector of the j-th phrase;
text representing counseling questions +.>Attention weight of the j-th phrase in (a);
respectively a parameter matrix;
t represents a transpose;
Representing the length of the weighted code representation result of the phrase;
s22: pre-processed consultation problem text data based on attention weight of phraseAnd (3) weighting:
wherein:
text data representing pre-processed counseling questions +.>Is a result of the weighting process;
text representing counseling questions +.>Attention weight of the L-th phrase;
text representing counseling questions +.>The coding vector of the L phrase;
s23: generating advisory problem textSemantic vector +.>：
Wherein:
representing a parameter matrix used to generate the semantic vector of the question text.
5. The artificial intelligence based custom closed-loop rule engine management control method according to claim 1, wherein in step S4, similarity calculation is performed between the semantic vector after the association expansion of the candidate document text and the semantic vector after the feature expansion of the consultation question text, including:
similarity calculation is carried out on the semantic vector after the association expansion of the candidate document text and the semantic vector after the feature expansion of the consultation problem text, wherein the semantic vector after the association expansionSemantic vector after feature expansion->The similarity calculation formula of (2) is:
wherein:
representing post-associative expansion semantic vector->Semantic after feature expansion Vector->Similarity of (2);
representing the reordering of the vectors of each column in the semantic vector according to the descending order of the attention weights of the phrases;
representing the L1 norm.
6. The artificial intelligence based custom closed-loop rule engine management control method according to claim 5, wherein in step S4, corresponding candidate document text is selected according to the similarity calculation result to perform customer service answer extraction, comprising:
selecting a customer service answer template in a candidate document text corresponding to the semantic vector after the association expansion with the highest similarity as a customer service answer extraction result, and taking the customer service answer extraction result as a consultation question textIs a reply to the user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311548638.4A CN117271755B (en) | 2023-11-21 | 2023-11-21 | Custom closed-loop rule engine management control method based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311548638.4A CN117271755B (en) | 2023-11-21 | 2023-11-21 | Custom closed-loop rule engine management control method based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117271755A CN117271755A (en) | 2023-12-22 |
| CN117271755B true CN117271755B (en) | 2024-03-08 |
Family
ID=89206655
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311548638.4A Active CN117271755B (en) | 2023-11-21 | 2023-11-21 | Custom closed-loop rule engine management control method based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117271755B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032741A (en) * | 2019-05-06 | 2019-07-19 | 重庆理工大学 | One kind is based on the relevant pseudo- document creation method of semantic extension and maximal margin |
| CN112000782A (en) * | 2020-08-01 | 2020-11-27 | 国网河北省电力有限公司信息通信分公司 | Intelligent customer service question-answering system based on k-means clustering algorithm |
| CN112632256A (en) * | 2020-12-29 | 2021-04-09 | 平安科技（深圳）有限公司 | Information query method and device based on question-answering system, computer equipment and medium |
| CN116450916A (en) * | 2023-04-11 | 2023-07-18 | 平安科技（深圳）有限公司 | Information query method and device based on fixed-segment classification, electronic equipment and medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6721179B2 (en) * | 2016-10-05 | 2020-07-08 | 国立研究開発法人情報通信研究機構 | Causal relationship recognition device and computer program therefor |
| US11704499B2 (en) * | 2021-04-13 | 2023-07-18 | Microsoft Technology Licensing, Llc | Generating questions using a resource-efficient neural network |
-
2023
- 2023-11-21 CN CN202311548638.4A patent/CN117271755B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032741A (en) * | 2019-05-06 | 2019-07-19 | 重庆理工大学 | One kind is based on the relevant pseudo- document creation method of semantic extension and maximal margin |
| CN112000782A (en) * | 2020-08-01 | 2020-11-27 | 国网河北省电力有限公司信息通信分公司 | Intelligent customer service question-answering system based on k-means clustering algorithm |
| CN112632256A (en) * | 2020-12-29 | 2021-04-09 | 平安科技（深圳）有限公司 | Information query method and device based on question-answering system, computer equipment and medium |
| CN116450916A (en) * | 2023-04-11 | 2023-07-18 | 平安科技（深圳）有限公司 | Information query method and device based on fixed-segment classification, electronic equipment and medium |
Non-Patent Citations (2)
| Title |
|---|
| "A Universal Quaternion Hypergraph Network for Multimodal Video Question Answering";Zhicheng Guo等;《IEEE Transactions on Multimedia ( Volume: 25)》;第1-4页 * |
| "基于 Spatial-DCTHash 动态参数网络的视觉问答算法";孟祥申等;《中国科学:信息科学》;第1008-1022页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117271755A (en) | 2023-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11151177B2 (en) | Search method and apparatus based on artificial intelligence | |
| CN108717406B (en) | Text emotion analysis method and device and storage medium | |
| CN107679039B (en) | Method and device for determining statement intention | |
| CN110442711B (en) | Text intelligent cleaning method and device and computer readable storage medium | |
| CN110866098B (en) | Machine reading method and device based on transformer and lstm and readable storage medium | |
| CN103019407B (en) | Input method application method, automatic question answering processing method, electronic equipment and server | |
| CN113378970B (en) | Sentence similarity detection method and device, electronic equipment and storage medium | |
| CN111767375A (en) | Semantic recall method and device, computer equipment and storage medium | |
| JP2022169743A (en) | Information extraction method and device, electronic equipment, and storage medium | |
| CN112699686A (en) | Semantic understanding method, device, equipment and medium based on task type dialog system | |
| CN113987125A (en) | Text structured information extraction method based on neural network and related equipment thereof | |
| CN113821622B (en) | Answer retrieval method and device based on artificial intelligence, electronic equipment and medium | |
| CN110765765B (en) | Contract key term extraction method, device and storage medium based on artificial intelligence | |
| CN113360654B (en) | Text classification method, apparatus, electronic device and readable storage medium | |
| CN113627797A (en) | Image generation method and device for employee enrollment, computer equipment and storage medium | |
| CN117271755B (en) | Custom closed-loop rule engine management control method based on artificial intelligence | |
| CN116881446A (en) | Semantic classification method, device, equipment and storage medium thereof | |
| CN114492437B (en) | Keyword recognition method and device, electronic equipment and storage medium | |
| CN116450829A (en) | Medical text classification method, device, equipment and medium | |
| CN116127066A (en) | Text clustering method, text clustering device, electronic equipment and storage medium | |
| CN114398482A (en) | Dictionary construction method and device, electronic equipment and storage medium | |
| CN112560427A (en) | Problem expansion method, device, electronic equipment and medium | |
| CN111221942A (en) | Intelligent text conversation generation method and device and computer readable storage medium | |
| CN111680513B (en) | Feature information identification method and device and computer readable storage medium | |
| CN117390173B (en) | Massive resume screening method for semantic similarity matching |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |