CN117056496A - Intelligent customer service interaction data management method based on big data - Google Patents

Intelligent customer service interaction data management method based on big data Download PDF

Info

Publication number
CN117056496A
CN117056496A CN202311314625.0A CN202311314625A CN117056496A CN 117056496 A CN117056496 A CN 117056496A CN 202311314625 A CN202311314625 A CN 202311314625A CN 117056496 A CN117056496 A CN 117056496A
Authority
CN
China
Prior art keywords
data
character
periodicity
type
characters
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
CN202311314625.0A
Other languages
Chinese (zh)
Other versions
CN117056496B (en
Inventor
李边芳
耿晓娜
黄湘云
邓栋
王亮
高晓磊
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.)
Qingdao Haier Lexinyun Technology Co ltd
Original Assignee
Qingdao Haier Lexinyun 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 Qingdao Haier Lexinyun Technology Co ltd filed Critical Qingdao Haier Lexinyun Technology Co ltd
Priority to CN202311314625.0A priority Critical patent/CN117056496B/en
Publication of CN117056496A publication Critical patent/CN117056496A/en
Application granted granted Critical
Publication of CN117056496B publication Critical patent/CN117056496B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/22—Indexing; Data structures therefor; Storage structures
    • G06F16/2219—Large Object storage; Management thereof
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/205—Parsing
    • G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/01—Customer relationship services
    • 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 data processing, in particular to an intelligent customer service interaction data management method based on big data. The method comprises the following steps: acquiring historical dialogue data; the method comprises the steps of obtaining a plurality of data segments and marking the data segments of historical dialogue data, obtaining the index position of each character in the data segments, and obtaining the periodicity of the characters according to the index position of the characters in the data segments and the number of the characters in the data segments; converting the data segment into sentence vectors so as to obtain the similarity of the data segment; acquiring a periodic influence factor according to the similarity of the data segments; acquiring the weighted periodicity of the character according to the periodicity of the character and the periodicity influencing factor; and periodically acquiring the frequency of the character according to the weight of the character and performing data compression. The invention reduces storage time and space.

Description

Intelligent customer service interaction data management method based on big data
Technical Field
The invention relates to the technical field of data processing, in particular to an intelligent customer service interaction data management method based on big data.
Background
With the advent of the big data age, more and more enterprises began to utilize intelligent customer service systems based on big data to optimize services and promote user experience. The big data intelligent customer service system needs to process a large amount of interaction data in the operation process, including questions of users, system answers, dialogue records and the like. The data volume of the information is huge, and the storage and processing of the data becomes very difficult and time-consuming, so that the data needs to be compressed to reduce the storage space and time and the storage cost.
rANS coding is a compression algorithm that takes into account both compression rate and compression time. Traditional rANS codes construct a fixed cumulative distribution table by counting the frequency of various characters in the data, and compress the data through the fixed cumulative distribution table. The intelligent customer service interactive system is continuously provided with real-time data input, and a fixed accumulated distribution table can not achieve a good compression effect when compressing data streams input in real time.
Disclosure of Invention
In order to solve the technical problem of poor compression effect, the invention provides an intelligent customer service interaction data management method based on big data, which adopts the following technical scheme:
the invention provides an intelligent customer service interaction data management method based on big data, which comprises the following steps:
acquiring historical dialogue data;
segmenting historical dialogue data to obtain a plurality of data segments and marking, obtaining all character types and index positions of each character for each data segment, and obtaining the periodicity of each type of character according to the index positions of each type of character in the data segments and the number of the characters of the data segments;
converting the data segments into sentence vectors, and calculating the similarity of the data segments adjacent to the labels according to the sentence vectors; acquiring a periodic influence factor according to the similarity of all adjacent data segments; acquiring the weighted periodicity of each type of character according to the periodicity of each type of character and the periodicity influencing factor;
and periodically acquiring the frequency of each type of character according to the weighting of each type of character, and carrying out data compression according to the frequency of each type of character.
Preferably, the method for segmenting the historical dialogue data to obtain a plurality of data segments and labeling is as follows:
usingThe word segmentation divides the historical dialogue data according to periods, exclamation marks and question marks in the historical dialogue data to obtain a plurality of data segments, and the sequence of the data segments is from small to large.
Preferably, the method for obtaining the periodicity of each type of character according to the index position of each type of character in the data segment and the number of characters in the data segment comprises the following steps:in (1) the->Representing the number of characters of type i in the a-th data segment,/and>index position representing the ith character of the b-th class in the a-th data segment, +.>The number of characters representing the a-th data segment, is->Representing the number of characters of type i in the a-1 data field,/for the data field>Index position representing the ith character of the b-th class in the a-1 data segment,/for the data segment>The number of characters representing the a-1 st data segment, m represents the number of data segments into which the history dialogue data is divided,/for>Representing the periodicity of the i-th character.
Preferably, the method for converting the data segment into sentence vector and calculating the similarity of the data segments with adjacent labels according to the sentence vector comprises the following steps:
by means ofThe model converts each data segment into a sentence vector, and for two adjacent sentence vectors with the labels, the cosine similarity of the two sentence vectors is calculated and is used as the similarity of the adjacent data segments corresponding to the adjacent sentence vectors.
Preferably, the method for obtaining the periodic influence factor according to the similarity of all adjacent data segments comprises the following steps:
in (1) the->Representing the similarity of the a-1 st data segment and the a-th data segment, +.>Representing the similarity of the a-2 th data segment and the a-1 st data segment, m representing the number of data segments into which the historical dialog data is divided, +.>Is super-parameter (herba Cinchi Oleracei)>Is a periodic influencing factor.
Preferably, the method for obtaining the weighted periodicity of each type of character according to the periodicity of each type of character and the periodicity influencing factor comprises the following steps:
let the product of the periodicity of each class of characters and the periodicity influencing factor be the weighted periodicity of each class of characters.
Preferably, the method for periodically obtaining the frequency of each type of character according to the weight of each type of character comprises the following steps:
let the ratio of the weighted periodicity of each class of characters to the sum of the weighted periodicity of all classes of characters be the frequency of each class of characters.
Preferably, the method for performing data compression according to the frequency of each type of character comprises the following steps:
and constructing a distribution accumulation table according to the frequency of each type of characters and the types of all characters, and compressing the historical dialogue data by using rANS coding algorithm.
Preferably, the method for acquiring the historical dialogue data comprises the following steps:
and collecting dialogue information once every preset time, wherein the collected dialogue information is all dialogue information which is collected last time and is completed this time, and recording the dialogue information collected each time as historical dialogue data.
Preferably, the method for acquiring the index position of the character comprises the following steps:
the characters in the data segment are numbered in the order from front to back, with the number being the index position of the character.
The invention has the following beneficial effects: according to the invention, the data is segmented, the periodicity of each type of character is calculated according to the distribution condition of the character in each segment of data, and the weighted periodicity of each type of character is calculated according to the data similarity change condition between each segment, so that the influence of the context data in the interaction process of the user and the intelligent customer service is avoided. According to the weighted periodicity of each type of character, the cumulative distribution table is adaptively adjusted, so that good compression effect is achieved when real-time data is compressed by using the adjusted cumulative distribution table, and therefore, the storage space and time are reduced, and the purpose of saving storage cost is achieved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of an intelligent customer service interaction data management method based on big data according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of specific implementation, structure, characteristics and effects thereof of the intelligent customer service interaction data management method based on big data according to the invention with reference to the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the intelligent customer service interaction data management method based on big data.
An embodiment of an intelligent customer service interaction data management method based on big data comprises the following steps:
referring to fig. 1, a flowchart of a method for managing intelligent customer service interaction data based on big data according to an embodiment of the present invention is shown, and the method includes the following steps:
step S001, acquiring history dialogue data.
The intelligent customer service interaction data refers to interaction information between a user and an intelligent customer service system, and comprises questions, answers, dialogue records and the like of the user; these data can be used to train and improve the capabilities of the intelligent customer service system, to increase the accuracy in the intelligent customer service interaction process, to increase more accurate and efficient service, to collect all dialogue information between the user and the intelligent customer system in the history record, to record as history dialogue information, where the history dialogue information is collected once without a fixed time, and in this embodiment, once every 1 hour.
Step S002, the historical dialogue data are segmented to obtain a plurality of data segments and marked, all character types and index positions of each character are obtained for each data segment, and the periodicity of each type of character is obtained according to the index positions of each type of character in the data segment and the number of the characters of the data segment.
Intelligent customer service is often used to solve problems for users and to provide assistance. Because of the variety and complexity of questions posed by users, they often need to interact through rounds to get satisfactory answers. Intelligent customer service systems typically keep track of the context information of user problems during interactions. This means that the intelligent customer service can use the interaction information between them to better understand the user's intent and needs in different rounds of the user's questions. Because of this contextual relationship, the user has a high degree of autocorrelation with the intelligent customer service interaction data.
In the interaction process of the user and the intelligent customer service, the two are continuously carried out in a one-question and one-answer mode, when the questions answered by the intelligent customer service and the questions proposed by the user are different, the user continuously asks questions by adjusting the question asking questions, and then the intelligent customer service system answers the user by analyzing keywords when the user continuously asks questions. Although the user can continuously adjust the questions asked during the questioning process, the keywords of the answers required by the user are unchanged, namely the keywords can appear for a plurality of times during the questioning process of the user, and the appearance positions are approximately similar. These data have a high degree of contextual autocorrelation and a context model can be built to predict data acquired in real time.
Firstly, counting all character types in the acquired historical dialogue data to obtain a character type set.
In the user questioning process, the position of a certain type of character in a sentence is roughly determined, for example, the basic composition of a sentence is: the subject, the predicate and the object are arranged, wherein the appearance position of the subject is at the beginning of a sentence, the appearance position of the predicate is in the middle of the sentence, the appearance position of the object is at the end of the sentence, and the problem description regulated by the user is the subject, the object and the complement. The subjects, predicates and objects are keywords, and the positions of the subjects, the predicates and the objects in each sentence are approximately the same, so that the user and intelligent customer service interaction data are required to be processed in a segmentation mode, each sentence in the interaction process is separated, and the periodicity of each character is acquired by acquiring the positions of each character in each sentence.
Thus utilizingSegmenting historical dialogue data, segmenting the historical dialogue data, separating finished text according to punctuation marks (periods, exclamation marks and question marks), enabling each period to serve as a data segment, and marking the separated data segments according to sequence, wherein commas do not separate sentences, and the text is divided into two parts by a text segment>The word segmentation is an existing algorithm, and is not described in detail herein.
The total number of characters for each data segment in the historical dialog data and the index position of each character on the data segment in each data segment are obtained.
The periodicity of a class of characters is noted as the index position of such characters within different data segments, and the formula is as follows:
in (1) the->Representing the number of characters of type i in the a-th data segment,/and>indicating the index position of the b-th character of the i-th type in the a-th data segment,the number of characters representing the a-th data segment, is->Indicating the number of characters of the i-th type in the a-1 data segment,index position representing the ith character of the b-th class in the a-1 data segment,/for the data segment>The number of characters representing the a-1 st data segment, m represents the number of data segments into which the history dialogue data is divided,/for>Representing the periodicity of the i-th character.
Wherein,indicating when the a-th data segment has +.>This +.>The ith character is at +.>Mean value of index positions in individual data segments, < >>Indicate->The i-th character in the data section is in +.>Positions in the individual data segments;
i.e. representing the +.>The index position of the class character changes by the interval size; />Indicate->The average index position interval of class characters in all data segments, i.e. +.>Periodicity of class character in all data segments, the larger the value, the +.>The longer the periodicity of the class character in all data segments, the smaller the value, the +.>The shorter the periodicity of the class character in all data segments.
Thus, the periodicity of each type of character in each history dialogue data is obtained.
Step S003, converting the data segments into sentence vectors, and calculating the similarity of the data segments adjacent to the labels according to the sentence vectors; acquiring a periodic influence factor according to the similarity of all adjacent data segments; and acquiring the weighted periodicity of each type of character according to the periodicity of each type of character and the periodicity influencing factor.
In the process of interaction between a user and the intelligent customer service system, the interactive information of the intelligent customer service is affected periodically due to the previous dialogue content, the questioning sequence of the user and the like, for example, when the user asks the same question, the intelligent customer service answers the content which is not needed by the user, the user replaces the convincing words again, and when the questioning sequence or dialogue content of the user becomes more complex and detailed, the intelligent customer service system may need more characters to answer the question or provide help; this may result in the character becoming longer periodically, so the system requires more characters to respond to the user's demand. Along with the change of time sequence, the periodicity of each type of character can change along with the questioning of the user, so that the periodicity influence factor of each type of character needs to be calculated according to the similarity change condition of each data segment in the interaction data of the user and the intelligent customer service along with the change of time sequence.
For each data segment useThe model converts it into sentence vectors, +.>The model is the prior art, and is not described in detail herein, and the similarity of two adjacent data segments is calculated according to two adjacent sentence vectors, and the formula is as follows:
in (1) the->A sentence vector representing the a-th data segment,sentence vector representing the a-1 st data segment,/->A modulus of the sentence vector representing the a-th data segment,modulo representing sentence vector of the a-1 st data segment,/for the sentence vector>Representing the similarity of the a-1 st data segment and the a-th data segment.
Wherein the method comprises the steps ofThe larger the value of (2), the +.>Data segment and->The greater the similarity of the individual data segments, i.e.>Data segment and->The more similar the individual data segments are, the less the two are affected by periodicity; />The smaller the value of (2), the +.>Data segment and->The smaller the similarity of the data segments, i.e.>Data segment numberThe more dissimilar the individual data segments, the more dissimilar the two will have a greater impact on periodicity.
Each data segment in the historical dialogue data is arranged according to a time sequence, the similarity of the data is continuously changed, the more the data segment which is closer to the real-time data has a larger periodical influence on the real-time data, the more the segmented data which is farther from the real-time data has a smaller periodical influence on the real-time data, namely the smaller the periodical influence on the real-time data by the data segment with a smaller label is, the larger the periodical influence on the real-time data by the data segment with a larger label is, and the real-time data is the last data segment in all the data segments.
And calculating a periodicity influence factor according to the similarity of adjacent data segments, wherein the formula is as follows:
in (1) the->Representing the similarity of the a-1 st data segment and the a-th data segment, +.>Representing the similarity of the a-2 th data segment and the a-1 st data segment, m representing the number of data segments into which the historical dialog data is divided, +.>Takes experience value as super parameter>，/>Is a periodic influencing factor of historical dialogue data.
Wherein the method comprises the steps ofThe similarity ratio of the similarity of the data segment at the back and the similarity of the data segment at the front is shown, and the larger the similarity ratio is, the higher the similarity of the data segment at the back is, the more the real-time data and the same character are more, and the character periodicity is smaller; the smaller the value, the higher the similarity of the data segment which is higher and is closer to the real-time data, the fewer the same characters appear, and the larger the character periodicity. />A periodic influencing factor representing all classes of characters, if->It is explained that the periodicity of all class characters becomes larger if +.>The periodicity of all class characters is reduced, if +.>The periodicity of all class characters is explained as unchanged. />Time sequence influencing parameter representing data segment, +.>The larger the data segment, the more real-time data the larger its corresponding timing impact parameter.
According to the periodicity of each type of character of the historical dialogue data and the weighted periodicity of each type of character in the periodic influence factor acquisition, the formula is as follows:
in (1) the->Is a periodic influencing factor, ++>Representing the periodicity of the i-th character, +.>Representing the weighted periodicity of the i-th character. The weighted periodicity of the i-th character is the prediction periodicity of the character of the next historical dialogue data.
Thus, the weighted periodicity of each type of character is obtained.
Step S004, the frequency of each type of character is obtained according to the weighted periodicity of each type of character, and data compression is carried out according to the frequency of each type of character.
The frequency of each type of character is periodically obtained according to the weight of each type of character obtained in the steps, and the formula is as follows:
in (1) the->Representing the weighted periodicity of the i-th character, n representing the number of character types,indicating the frequency of the i-th character.
Constructing the size according to the type of the characters and the frequency of each type of the charactersEach time information is collected corresponds to one historical dialogue data, the frequency of characters corresponding to the historical dialogue data is obtained according to the last collected historical dialogue data, the existing algorithm rANS codes are used for compressing according to the frequency, the compressed data are stored in the intelligent customer service interactive system so as to analyze the intelligent customer service interactive data subsequently, the intelligent customer service language model is continuously trained according to the intelligent customer service interactive data for each compressed data, and the language model is improved and optimized gradually along with the time.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.

Claims (10)

1. The intelligent customer service interaction data management method based on big data is characterized by comprising the following steps of:
acquiring historical dialogue data;
segmenting historical dialogue data to obtain a plurality of data segments and marking, obtaining all character types and index positions of each character for each data segment, and obtaining the periodicity of each type of character according to the index positions of each type of character in the data segments and the number of the characters of the data segments;
converting the data segments into sentence vectors, and calculating the similarity of the data segments adjacent to the labels according to the sentence vectors; acquiring a periodic influence factor according to the similarity of all adjacent data segments; acquiring the weighted periodicity of each type of character according to the periodicity of each type of character and the periodicity influencing factor;
and periodically acquiring the frequency of each type of character according to the weighting of each type of character, and carrying out data compression according to the frequency of each type of character.
2. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for segmenting the historical dialogue data to obtain a plurality of data segments and labeling is as follows:
usingThe word segmentation divides the historical dialogue data according to periods, exclamation marks and question marks in the historical dialogue data to obtain a plurality of data segments, and the sequence of the data segments is from small to large.
3. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for obtaining the periodicity of each type of character according to the index position of each type of character in the data segment and the number of characters in the data segment is as follows:
in (1) the->Representing the number of characters of type i in the a-th data segment,/and>indicating the index position of the b-th character of the i-th type in the a-th data segment,the number of characters representing the a-th data segment, is->Indicating the number of characters of the i-th type in the a-1 data segment,index position representing the ith character of the b-th class in the a-1 data segment,/for the data segment>The number of characters representing the a-1 st data segment, m represents the number of data segments into which the history dialogue data is divided,/for>Representing the periodicity of the i-th character.
4. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for converting the data segments into sentence vectors and calculating the similarity of the data segments with adjacent labels according to the sentence vectors is as follows:
by means ofThe model converts each data segment into a sentence vector, and for two adjacent sentence vectors with the labels, the cosine similarity of the two sentence vectors is calculated and is used as the similarity of the adjacent data segments corresponding to the adjacent sentence vectors.
5. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for obtaining the periodic influence factor according to the similarity of all adjacent data segments is as follows:
in (1) the->Representing the similarity of the a-1 st data segment and the a-th data segment, +.>Representing the similarity of the a-2 th data segment and the a-1 st data segment, m representing the number of data segments into which the historical dialog data is divided, +.>Is super-parameter (herba Cinchi Oleracei)>Is a periodic influencing factor.
6. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for obtaining the weighted periodicity of each type of character according to the periodicity of each type of character and the periodicity influencing factor is as follows:
let the product of the periodicity of each class of characters and the periodicity influencing factor be the weighted periodicity of each class of characters.
7. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for periodically obtaining the frequency of each type of character according to the weight of each type of character is as follows:
let the ratio of the weighted periodicity of each class of characters to the sum of the weighted periodicity of all classes of characters be the frequency of each class of characters.
8. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for data compression according to the frequency of each type of character is as follows:
and constructing a distribution accumulation table according to the frequency of each type of characters and the types of all characters, and compressing the historical dialogue data by using rANS coding algorithm.
9. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for acquiring the historical dialogue data is as follows:
and collecting dialogue information once every preset time, wherein the collected dialogue information is all dialogue information which is collected last time and is completed this time, and recording the dialogue information collected each time as historical dialogue data.
10. The intelligent customer service interaction data management method based on big data as claimed in claim 1, wherein the method for obtaining the index position of the character is as follows:
the characters in the data segment are numbered in the order from front to back, with the number being the index position of the character.
CN202311314625.0A 2023-10-12 2023-10-12 Intelligent customer service interaction data management method based on big data Active CN117056496B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311314625.0A CN117056496B (en) 2023-10-12 2023-10-12 Intelligent customer service interaction data management method based on big data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311314625.0A CN117056496B (en) 2023-10-12 2023-10-12 Intelligent customer service interaction data management method based on big data

Publications (2)

Publication Number Publication Date
CN117056496A true CN117056496A (en) 2023-11-14
CN117056496B CN117056496B (en) 2024-01-26

Family

ID=88664828

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311314625.0A Active CN117056496B (en) 2023-10-12 2023-10-12 Intelligent customer service interaction data management method based on big data

Country Status (1)

Country Link
CN (1) CN117056496B (en)

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE2307709A1 (en) * 1972-02-16 1973-08-23 Seismograph Service Corp MACHINE PROCESS FOR REDUCING SEISMIC DATA OBTAINED FROM A SEISMOGRAM GENERATED BY ARTIFICIAL EARTHQUAKE WAVES
US20030179114A1 (en) * 2002-03-21 2003-09-25 International Business Machines Corporation Method and system for improving lossless compression efficiency
JP2013143066A (en) * 2012-01-12 2013-07-22 Kddi Corp Question and answer program, server and method which use large amount of comment texts
US20200014399A1 (en) * 2015-10-01 2020-01-09 PacByte Solutions Pty Ltd Method and system for compressing and/or encrypting data files
US20200302316A1 (en) * 2019-03-18 2020-09-24 International Business Machines Corporation Question answering system influenced by user behavior and text metadata generation
CN115310429A (en) * 2022-08-05 2022-11-08 厦门靠谱云股份有限公司 Data compression and high-performance calculation method in multi-turn listening dialogue model
CN116192971A (en) * 2023-04-24 2023-05-30 江阴市华明电力发展集团有限公司 Intelligent cloud energy operation and maintenance service platform data management method
CN116521893A (en) * 2023-04-28 2023-08-01 苏州浪潮智能科技有限公司 Control method and control device of intelligent dialogue system and electronic equipment
CN116610265A (en) * 2023-07-14 2023-08-18 济南玖通志恒信息技术有限公司 Data storage method of business information consultation system
CN116722876A (en) * 2023-08-08 2023-09-08 北京点聚信息技术有限公司 Intelligent storage method for user data for format light reading
CN116827350A (en) * 2023-08-29 2023-09-29 众科云(北京)科技有限公司 Flexible work platform intelligent supervision method and system based on cloud edge cooperation

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE2307709A1 (en) * 1972-02-16 1973-08-23 Seismograph Service Corp MACHINE PROCESS FOR REDUCING SEISMIC DATA OBTAINED FROM A SEISMOGRAM GENERATED BY ARTIFICIAL EARTHQUAKE WAVES
US20030179114A1 (en) * 2002-03-21 2003-09-25 International Business Machines Corporation Method and system for improving lossless compression efficiency
JP2013143066A (en) * 2012-01-12 2013-07-22 Kddi Corp Question and answer program, server and method which use large amount of comment texts
US20200014399A1 (en) * 2015-10-01 2020-01-09 PacByte Solutions Pty Ltd Method and system for compressing and/or encrypting data files
US20200302316A1 (en) * 2019-03-18 2020-09-24 International Business Machines Corporation Question answering system influenced by user behavior and text metadata generation
CN115310429A (en) * 2022-08-05 2022-11-08 厦门靠谱云股份有限公司 Data compression and high-performance calculation method in multi-turn listening dialogue model
CN116192971A (en) * 2023-04-24 2023-05-30 江阴市华明电力发展集团有限公司 Intelligent cloud energy operation and maintenance service platform data management method
CN116521893A (en) * 2023-04-28 2023-08-01 苏州浪潮智能科技有限公司 Control method and control device of intelligent dialogue system and electronic equipment
CN116610265A (en) * 2023-07-14 2023-08-18 济南玖通志恒信息技术有限公司 Data storage method of business information consultation system
CN116722876A (en) * 2023-08-08 2023-09-08 北京点聚信息技术有限公司 Intelligent storage method for user data for format light reading
CN116827350A (en) * 2023-08-29 2023-09-29 众科云(北京)科技有限公司 Flexible work platform intelligent supervision method and system based on cloud edge cooperation

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
YIXIU WANG ET AL: "Multi-sentence Question Segmentation and Compression for Question Answering", 《NATURAL LANGUAGE PROCESSING AND CHINESE COMPUTING》 *
王秀华;: "基于相似度分段的数据智能自适应压缩算法", 信息通信, no. 08 *
程飞;刘凯;丁文文;时欢;张百戬;: "测量报告数据的谱分析压缩算法", 西安电子科技大学学报, no. 04 *

Also Published As

Publication number Publication date
CN117056496B (en) 2024-01-26

Similar Documents

Publication Publication Date Title
US20230377312A1 (en) System and method for neural network orchestration
CN110491416A (en) It is a kind of based on the call voice sentiment analysis of LSTM and SAE and recognition methods
CN109460737A (en) A kind of multi-modal speech-emotion recognition method based on enhanced residual error neural network
CN110457432B (en) Interview scoring method, interview scoring device, interview scoring equipment and interview scoring storage medium
CN110134946B (en) Machine reading understanding method for complex data
CN110222184A (en) A kind of emotion information recognition methods of text and relevant apparatus
CN109933652A (en) Intelligent answer method, apparatus, computer equipment and storage medium
Zhong et al. A Lightweight Model Based on Separable Convolution for Speech Emotion Recognition.
CN109582952A (en) Poem generation method, device, computer equipment and medium
CN112000771A (en) Judicial public service-oriented sentence pair intelligent semantic matching method and device
CN111859938B (en) Electronic medical record entity relation extraction method based on position vector noise reduction and rich semantics
CN115393933A (en) Video face emotion recognition method based on frame attention mechanism
JP2003132088A (en) Time series data retrieval system
CN110990628A (en) Method for solving video question and answer by utilizing multi-granularity convolutional network self-attention context network mechanism
CN116110405A (en) Land-air conversation speaker identification method and equipment based on semi-supervised learning
CN117056496B (en) Intelligent customer service interaction data management method based on big data
WO2021051507A1 (en) Bot conversation generation method, device, readable storage medium, and bot
CN114239575B (en) Statement analysis model construction method, statement analysis method, device, medium and computing equipment
CN115985320A (en) Intelligent device control method and device, electronic device and storage medium
CN113268628B (en) Music emotion recognition method based on modularized weighted fusion neural network
CN109885668A (en) A kind of expansible field interactive system status tracking method and apparatus
CN114792117A (en) Training method and device of session classification model and session classification method and device
Jones et al. Different words, same song: advice for substantively interpreting duration models
CN114116976A (en) Genetic algorithm-based intention identification method and system
CN112015760A (en) Automatic question-answering method and device based on candidate answer set reordering and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant