[go: up one dir, main page]

CN119202201A - Delivery control system and method based on large model and sound cloning technology - Google Patents

Delivery control system and method based on large model and sound cloning technology Download PDF

Info

Publication number
CN119202201A
CN119202201A CN202411676601.4A CN202411676601A CN119202201A CN 119202201 A CN119202201 A CN 119202201A CN 202411676601 A CN202411676601 A CN 202411676601A CN 119202201 A CN119202201 A CN 119202201A
Authority
CN
China
Prior art keywords
text
delivery
word
task
channel
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.)
Granted
Application number
CN202411676601.4A
Other languages
Chinese (zh)
Other versions
CN119202201B (en
Inventor
李龙
商廉
史伟男
徐鹏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Xinshiyun Science And Technology Co ltd
Original Assignee
Jiangsu Xinshiyun Science And Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangsu Xinshiyun Science And Technology Co ltd filed Critical Jiangsu Xinshiyun Science And Technology Co ltd
Priority to CN202411676601.4A priority Critical patent/CN119202201B/en
Publication of CN119202201A publication Critical patent/CN119202201A/en
Application granted granted Critical
Publication of CN119202201B publication Critical patent/CN119202201B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G10MUSICAL INSTRUMENTS; ACOUSTICS
    • G10LSPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L13/00Speech synthesis; Text to speech systems
    • G10L13/02Methods for producing synthetic speech; Speech synthesisers
    • G10L13/027Concept to speech synthesisers; Generation of natural phrases from machine-based concepts
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33Querying
    • G06F16/332Query formulation
    • G06F16/3329Natural language query formulation
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33Querying
    • G06F16/3331Query processing
    • G06F16/334Query execution
    • G06F16/3343Query execution using phonetics
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/35Clustering; Classification
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/04Architecture, e.g. interconnection topology
    • G06N3/045Combinations of networks
    • G06N3/0455Auto-encoder networks; Encoder-decoder networks
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/08Learning methods

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Mathematical Physics (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Acoustics & Sound (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Machine Translation (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a delivery control system and a delivery control method based on a large model and a sound cloning technology, which relate to the technical field of text delivery, and the invention classifies texts to be delivered in history, generates templates of each text and constructs a template database; the method comprises the steps of inputting user data into a text template, automatically generating a delivery text, constructing a text generation large model, selecting a receiving channel with highest comprehensive efficiency as a delivery channel, processing the text according to the type of the delivery channel, delivering the text to a delivery target, generating an evidence chain by using the data in the delivery process, judging whether a delivery task is successful or not according to the generated evidence chain, carrying out secondary delivery by using the optimized delivery channel, carrying out secondary judgment on the delivery task, communicating with the delivery target by using personalized voice, explaining text content to the delivery target, and carrying out opinion feedback after the text content is clear to the delivery target.

Description

Delivery control system and method based on large model and sound cloning technology
Technical Field
The invention relates to the technical field of text delivery, in particular to a delivery control system and method based on a large model and a sound cloning technology.
Background
Traditional legal text delivery relies primarily on manual delivery, mailing, or advertisement delivery. Legal staff needs to send text to the party by going to the gate or mailing. This approach is time consuming, costly, and difficult to guarantee delivery to power. With the rapid development of information technology, legal text delivery mode has undergone the transition from traditional manual delivery to modern intelligent and automatic delivery, especially the introduction of information technology, big data, artificial Intelligence (AI) and other technologies in recent years, so that legal text delivery control becomes more efficient, safe and standard. With the popularization of mobile phones, short message delivery gradually becomes a high-efficiency and convenient delivery mode, and is particularly suitable for short legal texts such as summons, reminding notices and the like. When it is difficult for a party to contact or avoid delivery, the system may post legal text on a designated website via a web advertisement, thereby completing the delivery procedure. However, with the development of informatization and the proliferation of delivery channels, how to select a proper delivery channel when delivering texts is difficult to judge in real time, and many delivery targets are not understood for most legal texts, so how to improve the participation of the delivery targets, understand texts more easily, and increase the text delivery efficiency is important.
Disclosure of Invention
The invention aims to provide a delivery control system and a delivery control method based on a large model and a sound cloning technology, so as to solve the problems in the prior art.
In order to achieve the above purpose, the present invention provides the following technical solutions:
the method for controlling the delivery based on the large model and the sound cloning technology comprises the following steps:
s100, collecting all texts to be sent in the history work, extracting the characteristics of each text, classifying the texts to be sent in the history, analyzing each type of text structure, generating templates of each text and constructing a template database;
Further, the specific steps of generating templates of each text and constructing a template database are as follows:
s101, collecting all texts to be sent in history work, and extracting words in each text in history as ,Representing the 1 st, 2 nd, 3 rd, n-th words in each text in the extracted history, wherein n is a positive integer, and calculating the number of each extracted word as follows,
Representing the number of times of occurrence of the 1 st, 2 nd, 3 rd, n th words in the calculated text, comparing the number of times of occurrence of each word in the text, taking the word with the largest number of occurrence as a text feature, and screening to obtain the feature of each text in the history, wherein the feature is as follows,The method comprises the steps of representing the characteristics of the 1 st, 2 nd, 3 rd, third and n th texts in the collected history, judging all texts in the collected history, classifying the texts with the same characteristics into the same category, and classifying the texts to obtain all the text categories as,The 1 st, 2 nd, 3 rd, m th texts obtained by classification, m being a positive integer;
s102, after classifying texts in the history, scanning each text image by using a scanning technology, and extracting the positions of all words in each text as ,The method comprises the steps of representing the positions of 1,2, 3, and p words in an extracted text, wherein p is a positive integer, analyzing the position of each word in the text, judging that adjacent words are the same part when each word is adjacent transversely, judging that the words are independent parts when the word positions are not adjacent words, and judging all words in the text to obtain that all word parts in the text are parts through judging all words in the text,Representing the 1 st, 2 nd, 3 rd, q th word portions obtained after text analysis, q being a positive integer;
S103, after analyzing all word parts in the text, extracting the word positions of the text of the same kind for h times, judging the extracted word positions, judging the corresponding words as fixed words when the positions of the same words in the h times of extraction are the same, judging the corresponding words as replacement words when the positions of the words in the h times of extraction are different, and judging the fixed words in each word part in the text as fixed words after the judgment ,Representing that the 1 st, 2 nd, 3 rd, i th fixed words in each word part are obtained through judgment, i is a positive integer, and the replacement words of each part in the text are obtained through judgment,The method comprises the steps of judging 1 st, 2 nd, 3 rd, j th replacement words in each word part, constructing a replacement word set of each text by utilizing all types of the replacement words, combining all word parts obtained by analyzing the text with corresponding fixed words to construct templates of each text, and storing all templates to obtain a template database.
S200, acquiring user data and text features to be generated of a target in real time, judging the types of the texts to be generated by using the text features, selecting templates corresponding to the types of the texts in a template database, inputting the user data into the text templates to automatically generate a text to be transmitted, and constructing a text generation large model;
The method comprises the steps of classifying text contents in history, generating templates of each text, training to obtain texts, automatically generating a large model, only extracting replacement words in user data to fill the large model, automatically generating required texts, replacing a large amount of manual writing work, saving time and labor cost, and automatically generating texts to avoid wrongly written words, grammar errors or logic inconsistencies possibly occurring in manual writing and ensure consistency and accuracy of output. In an information-intensive scenario, the automated generation system can quickly summarize data and information, generating meaningful reports.
Further, the specific steps of constructing the text to generate the large model are as follows:
S201, acquiring characteristics of a text to be generated in real time as Ts, searching a text template with the same characteristics in a template database by utilizing the acquired real-time text characteristics, and extracting a replacement word set in the text template after searching the text template with the characteristics of the real-time text;
S202, acquiring user data of a target in real time as ,Extracting the same kind of data from the collected user data by using the kind of words in the extracted replacement word set to obtain the input word of the generated text, extracting the position of each replacement word in the S103, marking the input word by using the word part position of the replacement word in the text, and setting the marked input word as the positive integer,Represents the 1 st, 2 nd, 3 rd, g th input data after marking,B represents the word part of the input data in the text, W represents the specific position of the input data in the word part of the text, and the marked input data is used for processing the text according to the word part of the input dataThe position of the text template is input into the searched text template, and a real-time delivery text is automatically generated;
s203, training the real-time characteristic search of the classification of the S201 and the generation of the text of the S202 by using machine learning, obtaining a text generation large model after training, and automatically generating the text which needs to be generated in real time in work by using the text generation large model.
S300, collecting all information receiving channels used by a delivery target in a history, calculating the comprehensive efficiency of each receiving channel in the history of the delivery target, selecting the receiving channel with the highest comprehensive efficiency as a conveying channel, processing a text according to the type of the conveying channel, and conveying the text to the delivery target;
further, the specific steps of processing the text according to the type of the conveying channel and conveying the text to the conveying target are as follows:
S301, collecting all information receiving channels used by the target in the history as ,The 1 st, 2 nd, 3 rd, r information receiving channels used for sending the target in the collected history, r being a positive integer;
S302, extracting the use times, the successful reception times and the reception time of each information receiving channel from the collected use data of each information receiving channel, and calculating the comprehensive efficiency of each information receiving communication, wherein the formula is as follows:
;
In the formula, ef represents the calculated comprehensive efficiency of each information receiving channel, C represents the number of times each information receiving channel is used, Representing total usage times of all information receiving channels, su represents successful times of reception after each information receiving channel is used, tt represents receiving time of each information receiving channel, and comprehensive efficiency of each information receiving channel is calculated as,The comprehensive efficiency of the r information receiving channels is calculated, and the information receiving channel with the highest comprehensive efficiency is compared and selected as a delivery channel for delivering the text;
s303, after the delivery channel of the delivered text is selected, carrying out data processing on the delivered text according to the transmission type of the delivery channel, and carrying out delivery on the delivered text after converting the delivered text into the information type of delivery.
Predicting the most effective delivery mode and time according to the behavior preference, the historical response time and the position of the delivery target, thereby improving the delivery power and efficiency;
S400, recording all data in the text delivery process in real time, generating an evidence chain by using the data in the delivery process, and judging whether the delivery task is successful or not according to the generated evidence chain;
Further, the specific steps of judging whether the delivery task is successful or not according to the generated evidence chain are as follows:
s401, analyzing the whole delivery process of the delivered text, recording each time the delivered text is processed, analyzing nodes, and constructing a delivery task chain as a text generation-channel selection-text conversion-transmission-delivery node -Receiving a feedback of the presence of the user,Representing each node reached during the text delivery,Representing delivery to a target node;
S402, when a delivery task is executed on a delivery text, recording the time of each step in a delivery task chain, marking the delivery task chain by using a time stamp to generate an evidence chain, collecting the time Tj when an information receiving channel selected in a history successfully receives information, and calculating an information delivery time threshold, wherein the formula is as follows: In the formula, ty represents a calculated time threshold value of the information receiving channel, Representing the time average of the information received by the selected information receiving channel in the collected history when the information was successfully received,Representing a time standard deviation of the information receiving channel selected from the collected history when the information receiving channel successfully receives the information;
S403, when delivering the text, extracting the time stamp of the sending stage in the evidence chain as Tf, collecting the real-time as Tm, judging the delivering task by using the information delivering time threshold, and when When feedback is not received, judging that the delivery task fails, and delivering the target to receive textAnd when feedback is received, judging that the delivery task is successful, and receiving the text by the delivery target.
The system is internally provided with a whole-course tracking function, an evidence chain of a delivery process is automatically generated, the evidence chain comprises time of each voice delivery, feedback of a receiver, recording of voice content and switching of delivery modes, delivery traceability is ensured, and all delivery behaviors and feedback information are recorded safely and tamper-proof.
S500, carrying out failure tracing on the delivery task by utilizing an evidence chain after judging that the delivery task fails, carrying out omnibearing optimization on a text delivery channel according to a tracing result, carrying out secondary delivery by utilizing the optimized delivery channel, and carrying out secondary judgment on the delivery task;
Further, the specific steps of carrying out secondary judgment on the transport task are as follows:
S501, after judging that a delivery task fails, extracting an evidence chain in the delivery process, checking the evidence chain for tracing, extracting delivery nodes in the evidence chain, and taking the last recorded delivery node as a delivery failure source point;
S502, screening according to the order of the comprehensive efficiency from high to low in S302, checking the delivery nodes needing to pass through in each information receiving channel, extracting the information receiving channels without delivery failure source points in the delivery channels, and selecting the delivery channel with the highest comprehensive efficiency in the extracted information receiving channels as the optimized delivery channel;
s503, conveying the delivery text by using the optimized conveying channel, and performing secondary judgment on the optimized delivery task according to the method in S403.
The method has the advantages that the source tracing searching can be clearly carried out on the failure reasons by utilizing the delivery nodes in the evidence chain to obtain failure source points, the delivery channel is optimized by utilizing the failure source points, the delivery of a new round is carried out, the reliability and the efficiency of the delivery are greatly enhanced, and the failure reason checking time is shortened.
And S600, when judging that the delivery task is completed and the text delivery is successful, the delivery target reads the delivery texts of different types in different modes by using the client, the client generates personalized voice according to the text types by using a sound cloning technology, communicates with the delivery target by using the personalized voice, interprets the text content to the delivery target, and feeds back comments after delivering the clear text content to the delivery target.
Further, the specific steps of feeding back after the target clear text content is sent are as follows:
S601, when judging that a delivery task is completed and text delivery is successful, a client acquires the type of the delivered text in real time, reads the delivered text in a corresponding type reading mode, generates personalized voice according to the text type by utilizing a sound cloning technology after reading the text, and utilizes the personalized voice to communicate with a delivery target, and utilizes a transformer algorithm to process problems raised by a user in the communication process, and the personalized voice solves the problems raised by the user;
S602, when the user understands the text content and responds and executes according to the text content, user opinion feedback is carried out through the client side, and the delivery task is completed.
When the system transmits a notice through voice, a user can perform real-time interactive inquiry through voice or text input to know text details or supplementary relevant information. The multi-mode interaction mode can enable the delivery process to be more efficient and flexible, and enhance the participation of the parties. Different types of voices are generated according to the types of the texts by utilizing a voice cloning technology, serious voices are generated when the texts are important and urgent, and relaxed voices are generated when the texts are non-urgent, so that the delivery target can feel the text contents in a self-cutting manner, and the delivery target can quickly understand the text contents.
The delivery control system based on the large model and the sound cloning technology comprises a data collection module, a text generation module, a channel selection module, a task judgment module, an optimization module and a personalized voice module;
The data collection module is used for collecting information receiving channels and use data which are sent to a target in the history;
The text generation module is used for classifying texts by utilizing texts generated in historical work, constructing templates of each text, and automatically generating the texts by utilizing the text templates;
The channel selection module is used for analyzing according to the collected information receiving channels used for delivering the target in the history, calculating the comprehensive efficiency and selecting a delivery channel for delivering the text;
The task judging module is used for recording the time stamp generated in the delivering process to form an evidence chain, calculating an information receiving time threshold according to the historical receiving time, and judging whether the delivering task is successful or not by using the time threshold and the evidence chain;
The optimizing module is used for tracing the failure by utilizing an evidence chain after judging that the delivery task fails to acquire a delivery failure source point, analyzing the delivery failure source point and optimizing and re-delivering the delivery channel;
The personalized voice module is used for generating personalized voice according to the text type by utilizing a voice cloning technology after the text is sent, communicating with a sending target by utilizing the personalized voice, processing the problem posed by the user by utilizing a transformer algorithm in the communication process, and solving the problem posed by the user by utilizing the personalized voice
The text generation module comprises a classification unit, a template construction unit and a text generation large model unit;
The classifying unit is used for extracting the characteristics of all texts generated in the history and classifying the texts according to the characteristics of each text;
The template construction unit is used for scanning the texts, searching word parts, fixed words and replacement words of the texts, and constructing templates of each text;
The text generation large model unit is used for collecting user data sent to a target, extracting replacement words in the user data as input words, marking the positions of the input words, inputting the positions of the input words into the template to automatically generate a text, and generating the text by utilizing machine learning training to generate a large model.
The task judging module comprises a delivering task chain unit, an evidence chain unit, a threshold unit and a judging unit;
The delivery task chain unit is used for analyzing the delivery process and generating a delivery task chain;
The evidence chain unit is used for extracting the time of delivering each node in the task chain, generating a time stamp mark and constructing an evidence chain;
the threshold unit is used for calculating and obtaining an information delivery time threshold according to the time of receiving information by the delivery channel in the history;
the judging unit is used for extracting the time stamp and the real-time of the sending stage in the evidence chain and judging the sending task by utilizing the time threshold.
Compared with the prior art, the invention has the beneficial effects that:
1. the method and the device generate templates of each text after classifying the text content in the history, train the templates to obtain the text, automatically generate a large model, only extract the replacement words in the user data to fill the large model to automatically generate the required text, replace a large amount of manual writing work, save time and labor cost, and automatically generate the text to avoid wrongly written words, grammar errors or logic inconsistencies possibly occurring in manual writing and ensure the consistency and accuracy of output.
2. The invention automatically generates the evidence chain of the delivery process, which comprises the time of each voice delivery, the feedback of a receiver, the recording of voice content and the switching process of delivery modes, ensures the legality and traceability of delivery, and ensures the transparency and fairness of legal delivery when all delivery behaviors and feedback information are recorded safely and non-falsely as evidence in future litigation or dispute.
3. The user can perform real-time interactive query through voice or text input, and know text details or supplement related information. By utilizing the multi-mode interaction mode, the delivering process can be more efficient and flexible, and the participation of the parties is enhanced.
Drawings
FIG. 1 is a block diagram of a delivery control system based on a large model and acoustic cloning technique in accordance with the present invention;
FIG. 2 is a schematic diagram showing steps of a method for controlling delivery based on a large model and a sound cloning technique according to the present invention;
FIG. 3 is a schematic diagram of a text template of a delivery control method based on a large model and a sound cloning technique according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Embodiment as shown in fig. 1-3, the present invention provides a technical solution,
The method for controlling the delivery based on the large model and the sound cloning technology comprises the following steps:
s100, collecting all texts to be sent in the history work, extracting the characteristics of each text, classifying the texts to be sent in the history, analyzing each type of text structure, generating templates of each text and constructing a template database;
The specific steps of generating templates of each text and constructing a template database are as follows:
s101, collecting all texts to be sent in history work, and extracting words in each text in history as ,Representing the 1 st, 2 nd, 3 rd, n-th words in each text in the extracted history, wherein n is a positive integer, and calculating the number of each extracted word as follows,
Representing the number of times of occurrence of the 1 st, 2 nd, 3 rd, n th words in the calculated text, comparing the number of times of occurrence of each word in the text, taking the word with the largest number of occurrence as a text feature, and screening to obtain the feature of each text in the history, wherein the feature is as follows,The method comprises the steps of representing the characteristics of the 1 st, 2 nd, 3 rd, third and n th texts in the collected history, judging all texts in the collected history, classifying the texts with the same characteristics into the same category, and classifying the texts to obtain all the text categories as,The 1 st, 2 nd, 3 rd, m th texts obtained by classification, m being a positive integer;
s102, after classifying texts in the history, scanning each text image by using a scanning technology, and extracting the positions of all words in each text as ,The method comprises the steps of representing the positions of 1,2, 3, and p words in an extracted text, wherein p is a positive integer, analyzing the position of each word in the text, judging that adjacent words are the same part when each word is adjacent transversely, judging that the words are independent parts when the word positions are not adjacent words, and judging all words in the text to obtain that all word parts in the text are parts through judging all words in the text,Representing the 1 st, 2 nd, 3 rd, q th word portions obtained after text analysis, q being a positive integer;
S103, after analyzing all word parts in the text, extracting the word positions of the text of the same kind for h times, judging the extracted word positions, judging the corresponding words as fixed words when the positions of the same words in the h times of extraction are the same, judging the corresponding words as replacement words when the positions of the words in the h times of extraction are different, and judging the fixed words in each word part in the text as fixed words after the judgment ,Representing that the 1 st, 2 nd, 3 rd, i th fixed words in each word part are obtained through judgment, i is a positive integer, and the replacement words of each part in the text are obtained through judgment,The method comprises the steps of judging 1 st, 2 nd, 3 rd, j th replacement words in each word part, constructing a replacement word set of each text by utilizing all types of the replacement words, combining all word parts obtained by analyzing the text with corresponding fixed words to construct templates of each text, and storing all templates to obtain a template database.
S200, acquiring user data and text features to be generated of a target in real time, judging the types of the texts to be generated by using the text features, selecting templates corresponding to the types of the texts in a template database, inputting the user data into the text templates to automatically generate a text to be transmitted, and constructing a text generation large model;
The method comprises the steps of classifying text contents in history, generating templates of each text, training to obtain texts, automatically generating a large model, only extracting replacement words in user data to fill the large model, automatically generating required texts, replacing a large amount of manual writing work, saving time and labor cost, and automatically generating texts to avoid wrongly written words, grammar errors or logic inconsistencies possibly occurring in manual writing and ensure consistency and accuracy of output. In an information-intensive scenario, the automated generation system can quickly summarize data and information, generating meaningful reports.
The specific steps of constructing the text generation large model are as follows:
S201, acquiring characteristics of a text to be generated in real time as Ts, searching a text template with the same characteristics in a template database by utilizing the acquired real-time text characteristics, and extracting a replacement word set in the text template after searching the text template with the characteristics of the real-time text;
S202, acquiring user data of a target in real time as ,Extracting the same kind of data from the collected user data by using the kind of words in the extracted replacement word set to obtain the input word of the generated text, extracting the position of each replacement word in the S103, marking the input word by using the word part position of the replacement word in the text, and setting the marked input word as the positive integer,Represents the 1 st, 2 nd, 3 rd, g th input data after marking,B represents the word part of the input data in the text, W represents the specific position of the input data in the word part of the text, and the marked input data is used for processing the text according to the word part of the input dataThe position of the text template is input into the searched text template, and a real-time delivery text is automatically generated;
s203, training the real-time characteristic search of the classification of the S201 and the generation of the text of the S202 by using machine learning, obtaining a text generation large model after training, and automatically generating the text which needs to be generated in real time in work by using the text generation large model.
S300, collecting all information receiving channels used by a delivery target in a history, calculating the comprehensive efficiency of each receiving channel in the history of the delivery target, selecting the receiving channel with the highest comprehensive efficiency as a conveying channel, processing a text according to the type of the conveying channel, and conveying the text to the delivery target;
the specific steps of processing the text according to the type of the conveying channel and conveying the text to the conveying target are as follows:
S301, collecting all information receiving channels used by the target in the history as ,The 1 st, 2 nd, 3 rd, r information receiving channels used for sending the target in the collected history, r being a positive integer;
S302, extracting the use times, the successful reception times and the reception time of each information receiving channel from the collected use data of each information receiving channel, and calculating the comprehensive efficiency of each information receiving communication, wherein the formula is as follows:
;
In the formula, ef represents the calculated comprehensive efficiency of each information receiving channel, C represents the number of times each information receiving channel is used, Representing total usage times of all information receiving channels, su represents successful times of reception after each information receiving channel is used, tt represents receiving time of each information receiving channel, and comprehensive efficiency of each information receiving channel is calculated as,The comprehensive efficiency of the r information receiving channels is calculated, and the information receiving channel with the highest comprehensive efficiency is compared and selected as a delivery channel for delivering the text;
s303, after the delivery channel of the delivered text is selected, carrying out data processing on the delivered text according to the transmission type of the delivery channel, and carrying out delivery on the delivered text after converting the delivered text into the information type of delivery.
Predicting the most effective delivery mode and time according to the behavior preference, the historical response time and the position of the delivery target, thereby improving the delivery power and efficiency;
S400, recording all data in the text delivery process in real time, generating an evidence chain by using the data in the delivery process, and judging whether the delivery task is successful or not according to the generated evidence chain;
the specific steps of judging whether the delivery task is successful or not according to the generated evidence chain are as follows:
s401, analyzing the whole delivery process of the delivered text, recording each time the delivered text is processed, analyzing nodes, and constructing a delivery task chain as a text generation-channel selection-text conversion-transmission-delivery node -Receiving a feedback of the presence of the user,Representing each node reached during the text delivery,Representing delivery to a target node;
S402, when a delivery task is executed on a delivery text, recording the time of each step in a delivery task chain, marking the delivery task chain by using a time stamp to generate an evidence chain, collecting the time Tj when an information receiving channel selected in a history successfully receives information, and calculating an information delivery time threshold, wherein the formula is as follows: In the formula, ty represents a calculated time threshold value of the information receiving channel, Representing the time average of the information received by the selected information receiving channel in the collected history when the information was successfully received,Representing a time standard deviation of the information receiving channel selected from the collected history when the information receiving channel successfully receives the information;
S403, when delivering the text, extracting the time stamp of the sending stage in the evidence chain as Tf, collecting the real-time as Tm, judging the delivering task by using the information delivering time threshold, and when When feedback is not received, judging that the delivery task fails, and delivering the target to receive textAnd when feedback is received, judging that the delivery task is successful, and receiving the text by the delivery target.
The system is internally provided with a whole-course tracking function, an evidence chain of a delivery process is automatically generated, the system comprises time of each voice delivery, feedback of a receiver, recording of voice content and switching of delivery modes, the validity and traceability of delivery are ensured, all delivery behaviors and feedback information are recorded safely and non-falsely and used as evidence in future litigation or dispute, and the transparency and fairness of legal delivery are ensured.
S500, carrying out failure tracing on the delivery task by utilizing an evidence chain after judging that the delivery task fails, carrying out omnibearing optimization on a text delivery channel according to a tracing result, carrying out secondary delivery by utilizing the optimized delivery channel, and carrying out secondary judgment on the delivery task;
The method for carrying out secondary judgment on the transport task comprises the following specific steps:
S501, after judging that a delivery task fails, extracting an evidence chain in the delivery process, checking the evidence chain for tracing, extracting delivery nodes in the evidence chain, and taking the last recorded delivery node as a delivery failure source point;
S502, screening according to the order of the comprehensive efficiency from high to low in S302, checking the delivery nodes needing to pass through in each information receiving channel, extracting the information receiving channels without delivery failure source points in the delivery channels, and selecting the delivery channel with the highest comprehensive efficiency in the extracted information receiving channels as the optimized delivery channel;
s503, conveying the delivery text by using the optimized conveying channel, and performing secondary judgment on the optimized delivery task according to the method in S403.
The method has the advantages that the source tracing searching can be clearly carried out on the failure reasons by utilizing the delivery nodes in the evidence chain to obtain failure source points, the delivery channel is optimized by utilizing the failure source points, the delivery of a new round is carried out, the reliability and the efficiency of the delivery are greatly enhanced, and the failure reason checking time is shortened.
And S600, when judging that the delivery task is completed and the text delivery is successful, the delivery target reads the delivery texts of different types in different modes by using the client, the client generates personalized voice according to the text types by using a sound cloning technology, communicates with the delivery target by using the personalized voice, interprets the text content to the delivery target, and feeds back comments after delivering the clear text content to the delivery target.
The specific steps of feeding back after the target clear text content is sent are as follows:
S601, when judging that a delivery task is completed and text delivery is successful, a client acquires the type of the delivered text in real time, reads the delivered text in a corresponding type reading mode, generates personalized voice according to the text type by utilizing a sound cloning technology after reading the text, and utilizes the personalized voice to communicate with a delivery target, and utilizes a transformer algorithm to process problems raised by a user in the communication process, and the personalized voice solves the problems raised by the user;
S602, when the user understands the text content and responds and executes according to the text content, user opinion feedback is carried out through the client side, and the delivery task is completed.
When the system transmits a notice through voice, a user can perform real-time interactive inquiry through voice or text input to know text details or supplementary relevant information. The multi-mode interaction mode can enable the delivery process to be more efficient and flexible, and enhance the participation of the parties. Different types of voices are generated according to the types of the texts by utilizing a voice cloning technology, serious voices are generated when the texts are important and urgent, and relaxed voices are generated when the texts are non-urgent, so that the delivery target can feel the text contents in a self-cutting manner, and the delivery target can quickly understand the text contents.
The delivery control system based on the large model and the sound cloning technology comprises a data collection module, a text generation module, a channel selection module, a task judgment module, an optimization module and a personalized voice module;
The data collection module is used for collecting information receiving channels and use data which are sent to a target in the history;
The text generation module is used for classifying texts by utilizing texts generated in historical work, constructing templates of each text, and automatically generating the texts by utilizing the text templates;
The channel selection module is used for analyzing according to the collected information receiving channels used for delivering the target in the history, calculating the comprehensive efficiency and selecting a delivery channel for delivering the text;
The task judging module is used for recording the time stamp generated in the delivering process to form an evidence chain, calculating an information receiving time threshold according to the historical receiving time, and judging whether the delivering task is successful or not by using the time threshold and the evidence chain;
The optimizing module is used for tracing the failure by utilizing an evidence chain after judging that the delivery task fails to acquire a delivery failure source point, analyzing the delivery failure source point and optimizing and re-delivering the delivery channel;
The personalized voice module is used for generating personalized voice according to the text type by utilizing a voice cloning technology after the text is sent, communicating with a sending target by utilizing the personalized voice, processing the problem posed by the user by utilizing a transformer algorithm in the communication process, and solving the problem posed by the user by utilizing the personalized voice
The text generation module comprises a classification unit, a template construction unit and a text generation large model unit;
The classifying unit is used for extracting the characteristics of all texts generated in the history and classifying the texts according to the characteristics of each text;
The template construction unit is used for scanning the texts, searching word parts, fixed words and replacement words of the texts, and constructing templates of each text;
The text generation large model unit is used for collecting user data sent to a target, extracting replacement words in the user data as input words, marking the positions of the input words, inputting the positions of the input words into the template to automatically generate a text, and generating the text by utilizing machine learning training to generate a large model.
The task judging module comprises a delivering task chain unit, an evidence chain unit, a threshold unit and a judging unit;
The delivery task chain unit is used for analyzing the delivery process and generating a delivery task chain;
The evidence chain unit is used for extracting the time of delivering each node in the task chain, generating a time stamp mark and constructing an evidence chain;
the threshold unit is used for calculating and obtaining an information delivery time threshold according to the time of receiving information by the delivery channel in the history;
the judging unit is used for extracting the time stamp and the real-time of the sending stage in the evidence chain and judging the sending task by utilizing the time threshold.
Analyzing legal texts generated in histories, classifying the texts into civil complaints, authorized delegated books and the like;
Using an authorized code book as an example, generating a text template by scanning and determining a word part, a fixed word and a replacement word as shown in fig. 3, wherein xxx represents the replacement word;
extracting and inputting replacement words in user data reaching a target, and automatically generating an authorization notice;
The information receiving channel used for the target in the history is a mailbox, an express, a short message and a telephone voice, and the mailbox with the highest comprehensive efficiency is obtained by calculation;
Calculating to obtain a delivery time threshold of 30min, extracting the sending time of 12:00 and the real-time of 13:00 in an evidence chain, judging that the number of times system is not fed back yet, and judging that delivery fails;
And tracing the source of failure by using the evidence chain to obtain the failure source point as the mailbox address error, reselecting the telephone voice as a conveying channel to carry out new conveying, and converting the text into voice data transmission.
After the delivery task is completed, the delivery target receives the delivery text, the text type is acquired by the client, the text type is voice, the client reads the authorization notice by adopting a voice reading mode, the voice cloning technology is utilized to generate personalized voice to explain the content of the authorization notice to the delivery target, the delivery target is helped to understand the text, the delivery target is understood, and user opinion feedback is carried out in the client according to the text when the delivery target is understood, so that delivery is completed.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.

Claims (10)

1.基于大模型及声音克隆技术的送达控制方法,其特征在于:所述方法包括以下步骤:1. A delivery control method based on a large model and sound cloning technology, characterized in that the method comprises the following steps: S100、收集历史工作中需要送达的所有文本,提取每个文本的特征,对历史中需要送达的文本进行分类,对每种类型的文本结构进行分析,生成每种文本的模板并构建模板数据库;S100, collecting all texts that need to be delivered in historical work, extracting features of each text, classifying the texts that need to be delivered in history, analyzing the structure of each type of text, generating templates for each type of text and building a template database; S200、实时采集送达目标的用户数据和需要生成的文本特征,利用文本特征判断需要生成的文本种类,并在模板数据库中选择对应文本种类的模板,将用户数据输入到文本模板中自动生成送达文本,构建文本生成大模型;S200, collecting user data of the delivery target and text features to be generated in real time, using the text features to determine the type of text to be generated, and selecting a template corresponding to the text type in the template database, inputting the user data into the text template to automatically generate the delivery text, and constructing a large model for text generation; S300、收集历史中送达目标使用的所有信息接收通道,计算送达目标历史中每种接收通道的综合效率,选择综合效率最高的接收通道作为运送通道,根据运送通道的类型对文本进行处理后向送达目标进行运送;S300, collecting all information receiving channels used by the delivery target in history, calculating the comprehensive efficiency of each receiving channel in the delivery target history, selecting the receiving channel with the highest comprehensive efficiency as the delivery channel, and processing the text according to the type of the delivery channel and delivering it to the delivery target; S400、实时记录文本送达过程中的所有数据,利用送达过程中的数据生成证据链,根据生成的证据链判断送达任务是否成功;S400, recording all data in the text delivery process in real time, using the data in the delivery process to generate an evidence chain, and judging whether the delivery task is successful based on the generated evidence chain; S500、当判断送达任务失败后,利用证据链对送达任务进行失败溯源,根据溯源结果对文本运送通道进行全方位优化,利用优化后的运送通道进行二次运送,并对运送任务进行二次判断;S500: When it is determined that the delivery task has failed, the evidence chain is used to trace the failure of the delivery task, and the text delivery channel is fully optimized according to the tracing result. The optimized delivery channel is used for secondary delivery, and the delivery task is re-judged; S600、当判断运送任务完成,文本运送成功后,送达目标利用客户端针对不同类型的送达文本采用不同方式进行读取,客户端利用声音克隆技术根据文本类型生成个性化语音,利用个性化语音和送达目标进行沟通,向送达目标解释文本内容,送达目标明确文本内容后进行意见反馈。S600. When it is determined that the delivery task is completed and the text is delivered successfully, the delivery target uses the client to read different types of delivery texts in different ways. The client uses sound cloning technology to generate personalized voice according to the text type, uses the personalized voice to communicate with the delivery target, explains the text content to the delivery target, and the delivery target provides feedback after clarifying the text content. 2.根据权利要求1所述的基于大模型及声音克隆技术的送达控制方法,其特征在于:所述S100中生成每种文本的模板并构建模板数据库的具体步骤为:2. The delivery control method based on large model and sound cloning technology according to claim 1 is characterized in that: the specific steps of generating a template for each text and constructing a template database in S100 are: S101、收集历史工作中需要送达的所有文本,提取历史中每个文本中的字词为表示提取的历史中每个文本中第1、2、3、...、n种字词,n为正整数;计算提取的每种字词的个数为S101, collect all the texts that need to be delivered in the historical work, and extract the words in each text in the history , Indicates the 1st, 2nd, 3rd, ..., nth word in each text in the extracted history, where n is a positive integer; the number of each word extracted is calculated as , 表示计算的文本中第1、2、3、...、n种字词出现的次数,对文本中的每种字词出现的次数进行对比,将出现次数最多的字词作为文本特征,筛查得到历史中每个文本的特征为表示收集的历史中第1、2、3、...、n个文本的特征,对收集的历史中所有文本的进行判断,将特征相同的文本归为同一种类,经过分类得到所有文本种类为表示分类得到的第1、2、3、...、m种文本,m为正整数; Indicates the number of occurrences of the 1st, 2nd, 3rd, ..., nth word in the calculated text. The number of occurrences of each word in the text is compared, and the word with the most occurrences is used as the text feature. The feature of each text in the history is screened as , Indicates the features of the 1st, 2nd, 3rd, ..., nth texts in the collected history. All the texts in the collected history are judged and the texts with the same features are classified into the same category. After classification, the categories of all texts are , represents the 1st, 2nd, 3rd, ..., mth texts obtained by classification, where m is a positive integer; S102、对历史中文本进行分类后,利用扫描技术对每种文本图像进行扫描,提取每种文本中所有字词的位置为表示提取的文本中第1、2、3、...、p个字词的位置,p为正整数;对文本中每个字词的位置进行分析,当每个字词在横向相邻时,判断相邻字词为同一部分;当字词位置没有相邻字词时,判断为独立部分;经过对文本中所有字词进行判断,得到文本中所有字词部分为表示对文本分析后得到的第1、2、3、...、q个字词部分,q为正整数;S102, after classifying the historical texts, use scanning technology to scan each text image and extract the positions of all words in each text. , Indicates the position of the 1st, 2nd, 3rd, ..., pth word in the extracted text, where p is a positive integer; analyze the position of each word in the text, and when each word is adjacent in the horizontal direction, the adjacent words are judged to be the same part; when there are no adjacent words in the word position, it is judged to be an independent part; after judging all the words in the text, the parts of all the words in the text are obtained. , represents the 1st, 2nd, 3rd, ..., qth word parts obtained after text analysis, where q is a positive integer; S103、当分析得到文本中的所有字词部分后,对相同种类文本的字词位置进行h次提取,对提取的字词位置进行判断,当在h次提取中相同字词的位置均相同时,判断对应字词为固定字词;当在h次提取中字词的位置不相同时,判断对应字词为替换字词;经过判断得到文本中每个字词部分中的固定字词为表示经过判断得到每个字词部分中第1、2、3、...、i个固定字词,i为正整数;判断得到文本中每个部分的替换字词为表示经过判断得到每个字词部分中第1、2、3、...、j个替换字词,j为正整数,利用所有替换字词种类构建每种文本的替换字词集合;将对文本分析得到所有字词部分和对应的固定字词进行结合构建得到每种文本的模板,之后将所有模板进行存储得到模板数据库。S103, after analyzing all word parts in the text, extract the word positions of the same type of text h times, and judge the extracted word positions. When the positions of the same words are the same in the h extractions, the corresponding words are judged to be fixed words; when the positions of the words are different in the h extractions, the corresponding words are judged to be replacement words; after judging, the fixed words in each word part in the text are , It means that the 1st, 2nd, 3rd, ..., i-th fixed words in each word part are obtained after judgment, where i is a positive integer; the replacement words for each part of the text are judged to be , It means that the 1st, 2nd, 3rd, ..., jth replacement words in each word part are obtained after judgment, j is a positive integer, and a replacement word set for each text is constructed using all replacement word types; all word parts and corresponding fixed words obtained by text analysis are combined to construct a template for each text, and then all templates are stored to obtain a template database. 3.根据权利要求2所述的基于大模型及声音克隆技术的送达控制方法,其特征在于:所述S200中构建文本生成大模型的具体步骤为:3. The delivery control method based on large model and sound cloning technology according to claim 2 is characterized in that: the specific steps of constructing the text generation large model in S200 are: S201、实时采集需要生成文本的特征为Ts,利用采集的实时文本特征在模板数据库中查找得到相同特征的文本模板,在查找得到需要实时文本特征的文本模板后,提取文本模板中替换字词集合;S201, collecting the features of the text to be generated in real time as Ts, using the collected real-time text features to search for a text template with the same features in a template database, and after finding the text template requiring the real-time text features, extracting a replacement word set in the text template; S202、实时采集送达目标的用户数据为表示采集的送达目标第1、2、3、...、v个用户数据,v为正整数;利用提取的替换字词集合中的字词种类在采集的用户数据中抽取相同种类的数据得到生成文本的输入字词,将所述S103中每个替换字词的位置提取,利用替换字词在文本中所在的字词部分位置对输入字词进行标记,设标记后的输入字词为表示进行标记后第1、2、3、...、g个输入数据,中B表示输入数据在文本中的字词部分,W表示输入数据在文本字词部分中的具体位置;将标记后的输入数据根据的位置输入到查找的文本模板中,自动生成实时送达文本;S202: Real-time collection of user data delivered to the target: , represents the first, second, third, ..., vth user data of the collected delivery target, where v is a positive integer; using the word types in the extracted replacement word set to extract the same type of data in the collected user data to obtain input words for generating text, extracting the position of each replacement word in S103, marking the input word using the word part position of the replacement word in the text, and setting the marked input word to , It indicates the 1st, 2nd, 3rd, ..., gth input data after marking. Where B represents the word part of the input data in the text, and W represents the specific position of the input data in the word part of the text; the marked input data is The location is input into the search text template to automatically generate real-time delivery text; S203、利用机器学习对S201分类实时特征查找和S202文本生成进行训练,经过训练后得到文本生成大模型,利用文本生成大模型对工作中实时需要生成的文本进行自动生成。S203. Use machine learning to train S201 classification real-time feature search and S202 text generation. After training, a large text generation model is obtained. The large text generation model is used to automatically generate text that needs to be generated in real time during work. 4.根据权利要求1所述的基于大模型及声音克隆技术的送达控制方法,其特征在于:所述S300中根据运送通道的类型对文本进行处理后向送达目标进行运送的具体步骤为:4. The delivery control method based on large model and sound cloning technology according to claim 1 is characterized in that: the specific steps of processing the text according to the type of delivery channel and then delivering it to the delivery target in S300 are: S301、收集历史中送达目标使用的所有信息接收通道为表示收集的历史中送达目标使用的第1、2、3、...、r种信息接收通道,r为正整数;对应收集历史中送达目标对每种信息接收通道的使用数据;S301, collect all the information receiving channels used by the delivery target in the history , Indicates the first, second, third, ..., rth information receiving channel used by the delivery target in the collected history, where r is a positive integer; corresponding to the usage data of each information receiving channel by the delivery target in the collected history; S302、在收集的每种信息接收通道的使用数据中提取每种信息接收通道的使用次数、接收成功次数和接收时间,计算每种信息接收通达的综合效率,公式为:S302, extracting the usage times, successful reception times and reception time of each information receiving channel from the collected usage data of each information receiving channel, and calculating the comprehensive efficiency of each information receiving channel, the formula is: ; 公式中,Ef表示计算的每种信息接收通道的综合效率,C表示提取的每种信息接收通道的使用次数,表示所有信息接收通道的总使用次数,Su表示每种信息接收通道使用后的接收成功次数,Tt表示每种信息接收通道的接收时间;经过计算得到每种信息接收通道的综合效率为表示计算的第1、2、3、...、r种信息接收通道的综合效率;对r个综合效率进行对比选择综合效率最大的信息接收通道作为送达文本的运送通道;In the formula, Ef represents the comprehensive efficiency of each information receiving channel calculated, C represents the number of times each information receiving channel is used. represents the total number of times all information receiving channels are used, Su represents the number of successful receptions after each information receiving channel is used, and Tt represents the receiving time of each information receiving channel. After calculation, the comprehensive efficiency of each information receiving channel is , Indicates the calculated comprehensive efficiency of the first, second, third, ..., r information receiving channels; compares the r comprehensive efficiencies and selects the information receiving channel with the largest comprehensive efficiency as the delivery channel for the delivered text; S303、在选择送达文本的运送通道后,根据运送通道的传输类型,对送达文本进行数据处理,转化为运送通达的信息类型后,将送达文本进行运送。S303. After selecting the delivery channel for the delivery text, data processing is performed on the delivery text according to the transmission type of the delivery channel, and after converting it into an information type suitable for delivery, the delivery text is delivered. 5.根据权利要求4所述的基于大模型及声音克隆技术的送达控制方法,其特征在于:所述S400中根据生成的证据链判断送达任务是否成功的具体步骤为:5. The delivery control method based on large model and sound cloning technology according to claim 4 is characterized in that: the specific steps of judging whether the delivery task is successful according to the generated evidence chain in S400 are: S401、对送达文本的整个送达过程进行分析,记录对送达文本每一次处理分析节点,构建送达任务链为文本生成-通道选择-文本转化-发送-送达节点-接收反馈,表示文本送达过程中到达的每一个节点,表示送达目标节点;S401: Analyze the entire delivery process of the delivery text, record each processing and analysis node of the delivery text, and build a delivery task chain of text generation-channel selection-text conversion-sending-delivery node - Receive feedback, Represents each node reached during the text delivery process, Indicates delivery to the target node; S402、对送达文本执行送达任务时,记录送达任务链中每一步的时间,利用时间戳对送达任务链进行标记生成证据链;收集历史中选择的信息接收通道成功接收信息时的时间Tj,计算信息送达时间阈值,公式为:,公式中,Ty表示计算的信息接收通道的时间阈值,表示收集的历史中选择的信息接收通道成功接收信息时的时间平均值,表示收集的历史中选择的信息接收通道成功接收信息时的时间标准差;S402, when executing the delivery task for the delivery text, record the time of each step in the delivery task chain, mark the delivery task chain with a timestamp to generate an evidence chain; collect the time Tj when the information receiving channel selected in the history successfully receives the information, and calculate the information delivery time threshold, the formula is: , where Ty represents the time threshold of the information receiving channel. It indicates the average time when the selected information receiving channel successfully receives information in the collected history. It represents the standard deviation of the time when the selected information receiving channel successfully receives the information in the collected history; S403、在送达文本进行运送时,提取证据链中发送阶段的时间戳为Tf,采集实时时间为Tm,利用信息送达时间阈值对送达任务进行判断,当还未接到反馈时,判断送达任务失败,送达目标为接收文本;当接收到反馈时,判断送达任务成功,送达目标接收到文本。S403, when the text is delivered, the timestamp of the sending stage in the evidence chain is extracted as Tf, the real-time collection time is Tm, and the delivery task is judged by the information delivery time threshold. If no feedback is received, the delivery task is considered to have failed and the delivery target is the received text. When feedback is received, it is determined that the delivery task is successful and the delivery target receives the text. 6.根据权利要求5所述的基于大模型及声音克隆技术的送达控制方法,其特征在于:所述S500中并对运送任务进行二次判断的具体步骤为:6. The delivery control method based on large model and sound cloning technology according to claim 5 is characterized in that: the specific steps of performing secondary judgment on the delivery task in S500 are: S501、当判断送达任务失败后,提取送达过程中的证据链,查看证据链进行溯源,提取证据链中的送达节点,将记录的最后一个送达节点作为送达失败源点;S501. When it is determined that the delivery task has failed, extract the evidence chain in the delivery process, check the evidence chain for source tracing, extract the delivery node in the evidence chain, and use the last delivery node recorded as the source point of the delivery failure; S502、在S302中根据综合效率从大到小的顺序进行筛选,对每个信息接收通道中需要经过的送达节点进行查看,提取出送达通道中没有送达失败源点信息接收通道,选择提取的信息接收通道中综合效率最大的作为优化后的运送通道;S502, screening in S302 in descending order of comprehensive efficiency, checking the delivery nodes that need to be passed in each information receiving channel, extracting the information receiving channels without the source of delivery failure in the delivery channels, and selecting the information receiving channels with the highest comprehensive efficiency among the extracted information receiving channels as the optimized delivery channel; S503、利用优化后的运送通道对送达文本进行运送,根据S403中的方法对优化的送达任务进行二次判断。S503, using the optimized delivery channel to deliver the delivery text, and performing a secondary judgment on the optimized delivery task according to the method in S403. 7.根据权利要求6所述的基于大模型及声音克隆技术的送达控制方法,其特征在于:所述S600中送达目标明确文本内容后进行反馈的具体步骤为:7. The delivery control method based on large model and voice cloning technology according to claim 6 is characterized in that: the specific steps of providing feedback after the delivery target specifies the text content in S600 are: S601、当判断运送任务完成,文本运送成功后,客户端实时采集送达文本的类型,采用对应类型的读取方式对送达文本进行读取,客户端在读取文本后利用声音克隆技术根据文本类型生成个性化语音,利用个性化语音和送达目标进行沟通,在沟通过程中利用transformer算法对用户提出的问题进行处理,个性化语音对用户提出的问题进行解答;S601, when it is determined that the delivery task is completed and the text is delivered successfully, the client collects the type of the delivered text in real time, and reads the delivered text using a reading method corresponding to the type. After reading the text, the client generates a personalized voice according to the text type using voice cloning technology, and communicates with the delivery target using the personalized voice. During the communication process, the transformer algorithm is used to process the questions raised by the user, and the personalized voice answers the questions raised by the user; S602、当用户理解文本内容,并根据文本内容进行响应执行时,通过客户端进行用户意见反馈,完成送达任务。S602: When the user understands the text content and responds according to the text content, user feedback is provided through the client to complete the delivery task. 8.基于大模型及声音克隆技术的送达控制系统,其特征在于:送达控制系统包括数据收集模块、文本生成模块、通道选择模块、任务判断模块、优化模块和个性化语音模块;8. A delivery control system based on a large model and voice cloning technology, characterized in that: the delivery control system includes a data collection module, a text generation module, a channel selection module, a task judgment module, an optimization module and a personalized voice module; 所述数据收集模块用于收集历史中送达目标使用的信息接收通道和使用数据;The data collection module is used to collect information receiving channels and usage data used by the delivery target in history; 所述文本生成模块用于利用历史工作中产生的文本,对文本进行分类后,构建每种文本的模板,利用文本模板自动生成文本;The text generation module is used to utilize the texts generated in the historical work, classify the texts, construct templates for each type of text, and automatically generate texts using the text templates; 所述通道选择模块用于根据收集的历史中送达目标使用的信息接收通道进行分析,计算综合效率,挑选出送达文本的运送通道;The channel selection module is used to analyze the information receiving channels used by the delivery target in the collected history, calculate the comprehensive efficiency, and select the delivery channel for the delivered text; 所述任务判断模块用于记录送达过程生成时间戳构成证据链,并根据历史接收时间计算信息接收时间阈值,利用时间阈值和证据链判断送达任务是否成功;The task judgment module is used to record the delivery process to generate a timestamp to form an evidence chain, and calculate the information reception time threshold based on the historical reception time, and use the time threshold and the evidence chain to determine whether the delivery task is successful; 所述优化模块用于在判断送达任务失败后,利用证据链对失败进行溯源,得到送达失败源点,对送达失败源点进行分析后,对运送通道进行优化重新运送;The optimization module is used to trace the failure by using the evidence chain after determining that the delivery task has failed, obtain the source of the delivery failure, analyze the source of the delivery failure, and optimize the delivery channel to re-deliver; 所述个性化语音模块用于在文本送达后客户端利用声音克隆技术根据文本类型生成个性化语音,利用个性化语音和送达目标进行沟通,在沟通过程中利用transformer算法对用户提出的问题进行处理,个性化语音对用户提出的问题进行解答。The personalized voice module is used to generate personalized voice according to the text type by the client using sound cloning technology after the text is delivered, communicate with the delivery target using the personalized voice, process the questions raised by the user using the transformer algorithm during the communication process, and answer the questions raised by the user with the personalized voice. 9.根据权利要求8所述的基于大模型及声音克隆技术的送达控制系统,其特征在于:所述文本生成模块包括分类单元、模板构建单元和文本生成大模型单元;9. The delivery control system based on large model and sound cloning technology according to claim 8, characterized in that: the text generation module includes a classification unit, a template construction unit and a text generation large model unit; 所述分类单元用于提取历史中产生的所有文本的特征,根据每个文本的特征对文本进行分类;The classification unit is used to extract the features of all texts generated in history and classify the texts according to the features of each text; 所述模板构建单元用于对文本进行扫描,查找得到文本的字词部分、固定字词和替换字词,构建每种文本的模板;The template construction unit is used to scan the text, find the word parts, fixed words and replacement words of the text, and construct a template for each text; 所述文本生成大模型单元用于对送达目标的用户数据进行收集,提取用户数据中的替换字词作为输入字词,对输入字词的位置进行标记,输入到模板中自动生成文本,并利用机器学习训练生成文本生成大模型。The text generation model unit is used to collect user data of the delivery target, extract replacement words in the user data as input words, mark the position of the input words, input them into the template to automatically generate text, and use machine learning training to generate a text generation model. 10.根据权利要求8所述的基于大模型及声音克隆技术的送达控制系统,其特征在于:所述任务判断模块包括送达任务链单元、证据链单元、阈值单元和判断单元;10. The delivery control system based on large model and sound cloning technology according to claim 8, characterized in that: the task judgment module includes a delivery task chain unit, an evidence chain unit, a threshold unit and a judgment unit; 所述送达任务链单元用于对送达过程进行分析,生成送达任务链;The delivery task chain unit is used to analyze the delivery process and generate a delivery task chain; 所述证据链单元用于对送达任务链中每个节点的时间进行提取,生成时间戳标记,构建证据链;The evidence chain unit is used to extract the time of delivery to each node in the task chain, generate a timestamp mark, and construct an evidence chain; 所述阈值单元用于根据历史中运送通道接收信息的时间计算得到信息送达时间阈值;The threshold unit is used to calculate the information delivery time threshold according to the time when the transport channel receives the information in history; 所述判断单元用于提取证据链中发送阶段的时间戳和实时时间,利用时间阈值对送达任务进行判断。The judgment unit is used to extract the timestamp and real time of the sending stage in the evidence chain, and use the time threshold to judge the delivery task.
CN202411676601.4A 2024-11-22 2024-11-22 Delivery control system and method based on large model and sound cloning technology Active CN119202201B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202411676601.4A CN119202201B (en) 2024-11-22 2024-11-22 Delivery control system and method based on large model and sound cloning technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202411676601.4A CN119202201B (en) 2024-11-22 2024-11-22 Delivery control system and method based on large model and sound cloning technology

Publications (2)

Publication Number Publication Date
CN119202201A true CN119202201A (en) 2024-12-27
CN119202201B CN119202201B (en) 2025-03-28

Family

ID=94070753

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202411676601.4A Active CN119202201B (en) 2024-11-22 2024-11-22 Delivery control system and method based on large model and sound cloning technology

Country Status (1)

Country Link
CN (1) CN119202201B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9785705B1 (en) * 2014-10-16 2017-10-10 Google Inc. Generating and applying data extraction templates
CN117541210A (en) * 2023-11-21 2024-02-09 上海道律信息技术有限公司 Intensive document delivery method and device

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9785705B1 (en) * 2014-10-16 2017-10-10 Google Inc. Generating and applying data extraction templates
CN117541210A (en) * 2023-11-21 2024-02-09 上海道律信息技术有限公司 Intensive document delivery method and device

Also Published As

Publication number Publication date
CN119202201B (en) 2025-03-28

Similar Documents

Publication Publication Date Title
CN111182162B (en) Telephone quality inspection method, device, equipment and storage medium based on artificial intelligence
CN118839003A (en) Large model-based examination question generation method
CN116521821A (en) Text semantic matching method and refrigeration equipment system
CN109460558B (en) Effect judging method of voice translation system
CN117992601A (en) Document generation method and device based on artificial intelligence
CN111985250A (en) Model training method, device and system and computer readable storage medium
CN114528894B (en) A training method for a follow-up question model and a follow-up question method for interview questions
CN119202201B (en) Delivery control system and method based on large model and sound cloning technology
CN111613107A (en) Artificial intelligence operating system
CN112948251B (en) Automatic software testing method and device
CN114782110A (en) Demand mining method and system based on logistic regression two-classification and JMTS
CN113642337B (en) Data processing method and device, translation method, electronic device, and computer-readable storage medium
CN120448266A (en) Non-functional testing method, device and electronic equipment for software system
CN118708706A (en) A method and device for assisting enterprise carbon emission accounting process based on multimodal large model
CN117193889A (en) Construction method of code example library and use method of code example library
CN111261167A (en) Automatic tag generation method for audio hot content
CN117037775A (en) Customer demand analysis method and system for incoming call
CN111563240B (en) Intelligent robot management system based on information resource library
CN111192008B (en) Self-help guiding system and method
CN118278898B (en) Electronic work order reporting method and system based on artificial intelligence semantic recognition
CN120725026B (en) A Knowledge Graph-Based Fragmented Semantic Understanding-Based Order Building System and Method
CN117708337B (en) Man-machine interaction method and system oriented to complex localization
CN118864191A (en) A student homework evaluation method and system based on large language model
CN113434946B (en) Building model checking system and method based on BIM
CN121070958A (en) Intelligent reply method based on feedback opinion and related device

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