CN116226356A - NLP-based intelligent customer service interaction method and system - Google Patents
NLP-based intelligent customer service interaction method and system Download PDFInfo
- Publication number
- CN116226356A CN116226356A CN202310507094.0A CN202310507094A CN116226356A CN 116226356 A CN116226356 A CN 116226356A CN 202310507094 A CN202310507094 A CN 202310507094A CN 116226356 A CN116226356 A CN 116226356A
- Authority
- CN
- China
- Prior art keywords
- dialogue
- interactive
- customer service
- interval
- keyword
- 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
Links
- 230000003993 interaction Effects 0.000 title claims abstract description 82
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000002452 interceptive effect Effects 0.000 claims abstract description 164
- 239000013598 vector Substances 0.000 claims abstract description 87
- 238000005457 optimization Methods 0.000 claims abstract description 27
- 238000004458 analytical method Methods 0.000 claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 13
- 230000011218 segmentation Effects 0.000 claims abstract description 9
- 238000012216 screening Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000000737 periodic effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 6
- 238000003058 natural language processing Methods 0.000 description 36
- 230000008569 process Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 6
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 230000004044 response Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000013075 data extraction Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/374—Thesaurus
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Development Economics (AREA)
- Artificial Intelligence (AREA)
- Accounting & Taxation (AREA)
- Human Computer Interaction (AREA)
- Economics (AREA)
- Finance (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Machine Translation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an intelligent customer service interaction method and system based on NLP, and relates to the technical field of data processing, wherein the method comprises the following steps: collecting an interactive dialogue data set of a target sample customer service; performing single-round interactive dialogue recognition on the interactive data in the interactive dialogue data set by taking the interactive object as a segmentation point; calling an interactive dialogue group in the customer service interactive corpus according to the interactive optimization instruction; performing feature analysis on the interactive dialogue group, and outputting interval features for identifying keyword vectors in two rounds of dialogue; carrying out keyword vector distance optimization according to the interval characteristics, and outputting a first optimal distance interval, wherein the first optimal distance interval is the maximum distance interval of the keyword vectors in two-round conversations; inputting the first optimal distance interval into an NLP recognition module, and optimizing feedback dialogue contents of target sample customer service. The invention solves the technical problems of low intelligent degree of intelligent customer service interaction and low service quality in the prior art, and achieves the technical effect of improving the interaction quality.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to an intelligent customer service interaction method and system based on NLP.
Background
Along with the continuous iterative updating of the technology, intelligent customer service with high feedback speed and high communication efficiency gradually replaces manual customer service with higher cost. Intelligent customer service can provide fine-grained knowledge management technology for enterprises, so that most enterprise users can obtain standardized services. Therefore, the research on the intelligent customer service interaction related technology has very important significance for improving the service quality of enterprises.
However, the accuracy of the problem reply content of the enterprise user by the current intelligent customer service is not enough, the technical processing error rate of the problem is higher, the problem raised by the user cannot be well replied, and the requirement of the user cannot be met. In the prior art, the intelligent customer service interaction degree is low, and the service quality is low.
Disclosure of Invention
The application provides an intelligent customer service interaction method and system based on NLP, which are used for solving the technical problems of low intelligent degree of intelligent customer service interaction and low service quality in the prior art.
In view of the above problems, the present application provides an intelligent customer service interaction method and system based on NLP.
In a first aspect of the present application, an intelligent customer service interaction method based on NLP is provided, wherein the method is applied to an intelligent customer service interaction system, the system includes an NLP recognition module and a word vector processing module, and the method includes:
collecting an interactive dialogue data set of a target sample customer service;
carrying out single-round interactive dialogue identification on the interactive data in the interactive dialogue data set by taking the interactive objects as segmentation points, and outputting interactive dialogue groups, wherein the interactive objects in each group of interactive dialogue at least comprise one user object and one customer service object;
storing the interactive dialogue group into a customer service interactive corpus, and calling the interactive dialogue group in the customer service interactive corpus when the intelligent customer service interactive system receives an interactive optimization instruction;
through carrying out feature analysis on the interactive dialogue groups in the customer service interactive corpus, outputting interval features for identifying keyword vectors in two rounds of dialogue;
performing keyword vector distance optimization according to the interval features, and outputting a first optimal distance interval, wherein the first optimal distance interval is a maximum distance interval of keyword vectors in two-round conversations;
and inputting the first optimal distance interval into the NLP recognition module, and optimizing the feedback dialogue content of the target sample customer service.
In a second aspect of the present application, there is provided an intelligent customer service interaction system based on NLP, the system comprising:
the data set acquisition module is used for acquiring an interactive dialogue data set of the target sample customer service;
the dialogue group output module is used for carrying out single-round interaction dialogue identification on the interaction data in the interaction dialogue data set by taking the interaction objects as the segmentation points and outputting interaction dialogue groups, and the interaction objects in each group of interaction dialogue at least comprise a user object and a customer service object;
the dialogue group calling module is used for storing the interactive dialogue group into a customer service interaction corpus, and calling the interactive dialogue group in the customer service interaction corpus when the intelligent customer service interaction system receives an interaction optimization instruction;
the interval feature output module is used for outputting interval features for identifying keyword vectors in two rounds of conversations by carrying out feature analysis on the interactive conversation groups in the customer service interactive corpus;
the first distance interval is a maximum distance interval of keyword vectors in two-wheel conversations;
and the dialogue content optimizing module is used for inputting the first distance interval into the NLP identifying module and optimizing the feedback dialogue content of the target sample customer service.
One or more technical solutions provided in the present application have at least the following technical effects or advantages:
according to the method, an interactive dialogue data set of target sample customer service is acquired, then single-round interactive dialogue identification is conducted on interactive data in the interactive dialogue data set by taking interactive objects as segmentation points, an interactive dialogue group is output, the interactive objects in each group of interactive dialogue at least comprise one user object and one customer service object, the interactive dialogue group is further stored in a customer service interaction corpus, when an intelligent customer service interaction system receives an interaction optimization instruction, the interactive dialogue group in the customer service interaction corpus is called, then feature analysis is conducted on the interactive dialogue group in the customer service interaction corpus, interval features for identifying keyword vectors in two rounds of dialogue are output, keyword vector distance optimization is conducted according to the interval features, a first optimal distance interval is output, the first optimal distance interval is a maximum distance interval of the keyword vectors in two rounds of dialogue, and then the first optimal distance interval is input into an NLP identification module to optimize feedback dialogue contents of the target sample customer service. The customer service interactive feedback content quality is improved, and the technical effect of improving the service quality is achieved by utilizing the NLP recognition module to intelligently optimize the content.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic flow chart of an intelligent customer service interaction method based on NLP according to an embodiment of the present application;
fig. 2 is a schematic flow chart of outputting a second reply dialogue in the intelligent customer service interaction method based on NLP according to the embodiment of the present application;
fig. 3 is a schematic flow chart of outputting a first distance interval in an intelligent customer service interaction method based on NLP according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of an intelligent customer service interaction system based on NLP according to an embodiment of the present application.
Reference numerals illustrate: the system comprises a data set acquisition module 11, a dialogue group output module 12, a dialogue group calling module 13, an interval characteristic output module 14, a distance interval output module 15 and a dialogue content optimization module 16.
Detailed Description
The application provides an intelligent customer service interaction method and system based on NLP, which are used for solving the technical problems of low intelligent degree and low service quality of intelligent customer service interaction in the prior art.
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present application based on the embodiments herein.
It should be noted that the terms "comprises" and "comprising," along with any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus, but may include other steps or modules not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
As shown in fig. 1, the present application provides an intelligent customer service interaction method based on NLP, wherein the method is applied to an intelligent customer service interaction system, the system includes an NLP recognition module and a word vector processing module, and the method includes:
step S100: collecting an interactive dialogue data set of a target sample customer service;
specifically, the intelligent customer service interaction system is a system for carrying out customer service inquiry and answer interaction based on an NLP technology, and through the system, the problems consulted by enterprise users can be intelligently replied. The intelligent customer service interaction system comprises an NLP recognition module and a word vector processing module. The NLP is a sub-field of artificial intelligence, namely a natural language processing field, and information is processed in a man-machine interaction mode. The NLP recognition module is a functional module for analyzing an input semantic scene and then performing intelligent answer of questions and answers. The word vector processing module is a functional module for carrying out vectorization processing on the keywords and sentences in the word stock and describing the keywords and sentences from a plurality of dimensions to obtain word vectors.
In one possible embodiment, the target sample customer service is any customer service that intelligently interacts with the questions consulted by the user based on an intelligent customer service interaction system. The interactive dialogue data set is a data set obtained after recording the dialogue of the interactive response process between the target sample customer service and the user, and comprises information such as interactive dialogue content, interactive dialogue objects and the like. And acquiring the interactive response content of the target sample customer service by acquiring the interactive dialogue data set, and providing basic data information for subsequent analysis.
Step S200: carrying out single-round interactive dialogue identification on the interactive data in the interactive dialogue data set by taking the interactive objects as segmentation points, and outputting interactive dialogue groups, wherein the interactive objects in each group of interactive dialogue at least comprise one user object and one customer service object;
step S300: storing the interactive dialogue group into a customer service interactive corpus, and calling the interactive dialogue group in the customer service interactive corpus when the intelligent customer service interactive system receives an interactive optimization instruction;
in one possible embodiment, the interactive dialogue group is obtained according to the recognition result by extracting the interactive object of the interactive data in the interactive dialogue data set and recognizing the single-round interactive dialogue in the data set by taking the interactive object as a segmentation point. Wherein the interactive dialogue group is provided with at least one single-round interactive dialogue. And each user and the target sample customer service in the interactive dialogue data set have interactive dialogue data, and the consultation problem of any user and the answer of the target sample customer service to the consultation problem are used as a round of interactive dialogue. Illustratively, the user asks "what is today weather? The target sample customer service replies "today in the cloudy day with 15-20 ℃ temperature", and the one-to-one reply is a single interactive dialogue.
Specifically, the interactive objects in each group of interactive dialogue at least comprise a user object and a customer service object, so that the occurrence of customer service of a user and a target sample in the interactive dialogue can be ensured, and the question-answering content in the follow-up analysis question-answering process is interactive data.
Specifically, the customer service interaction corpus is a database for storing interaction data for serving a target sample customer service. The interaction optimization instruction is a command which is issued by the intelligent customer service interaction system and needs to optimize the question and answer content. And after the intelligent customer service interaction system receives the interaction optimization instruction, the interaction dialogue group in the customer service interaction corpus is called by the system to provide basic data for the subsequent steps.
Step S400: through carrying out feature analysis on the interactive dialogue groups in the customer service interactive corpus, outputting interval features for identifying keyword vectors in two rounds of dialogue;
further, step S400 in the embodiment of the present application further includes:
step S410: taking a single-round dialogue in the interactive dialogue group as an optimizing unit, carrying out keyword vector interval feature recognition on the target sample customer service by using the interactive dialogue group called in the customer service interactive corpus, and outputting interval period features and interval frequency features;
step S420: and outputting the interval periodic characteristic and the interval frequency characteristic as the interval characteristic.
Specifically, feature analysis is performed on the interactive dialogue group in the customer service interactive corpus, and keyword vector features in the interactive dialogue group are extracted, so that the interval features for identifying two rounds of dialogue keyword vectors are obtained. The interval feature is a feature for describing the interval between keyword vectors in two rounds of conversations, and comprises an interval period feature and an interval frequency feature. The interval period feature is a feature for describing a distance period of occurrence of two rounds of dialogue keyword vectors. The interval frequency characteristic is a characteristic for describing the successful times of the two-round dialogue keyword vector comparison.
In one embodiment of the application, by randomly selecting a single-round dialogue from the interactive dialogue group as an optimizing unit and using the interactive dialogue group called in the customer service interactive corpus as an optimizing space, performing interval feature recognition on the single-round dialogue and the dialogues in the interactive dialogue group. The method comprises the steps of extracting keyword vectors in a single-round dialogue, comparing and analyzing the keyword vectors with keyword vectors of the dialogue in an interactive dialogue group to obtain distance periods and interval frequencies of the distance, and outputting interval period characteristics and interval frequency characteristics according to obtained results.
In one possible embodiment, by converting the interactive dialogue text in a single-round dialogue into word vectors, describing the meaning of the interactive dialogue by the word vectors, preferably, firstly converting the dialogue text into serial numbers without any meaning, and further upgrading the serial numbers into high-dimensional features with specific task properties by using ebedding, such as good weather, low temperature and the like, and obtaining the keyword vectors according to the high-dimensional features.
Step S500: performing keyword vector distance optimization according to the interval features, and outputting a first optimal distance interval, wherein the first optimal distance interval is a maximum distance interval of keyword vectors in two-round conversations;
further, as shown in fig. 3, step S500 in the embodiment of the present application further includes:
step S510: acquiring service environment information of the target sample customer service;
step S520: determining a keyword set according to the service environment information;
step S530: identifying the interactive dialogue group in the customer service interactive corpus based on the keyword set, and outputting keyword sentence vectors based on the keyword set;
step S540: and carrying out corpus positioning analysis according to the keyword sentence vector, and outputting a first distance interval.
Further, step S400 in the embodiment of the present application further includes:
step S541: invoking an interactive dialogue group in the customer service interactive corpus, and outputting a screening corpus based on the keyword set;
step S542: and carrying out word vectorization representation on the screening corpus, and outputting vectorization results, wherein the vectorization results comprise word vectors based on the keyword set.
Further, corpus positioning analysis is performed according to the keyword sentence vectors, a first distance interval is output, and the calculation formula of the keyword sentence vectors is as follows:
wherein ,
characterizing sentence vectors; m represents the number of keywords in the called interactive dialog group; />
A word vector characterizing each keyword; />
The weight of each keyword is characterized.
In one embodiment, the first preference interval is data describing a maximum distance interval of the keyword vectors in the two-round dialogue, including a maximum distance and a minimum distance of the keyword vectors in the two-round dialogue. The service environment information is information describing the background of the question-answer interaction of the target sample customer service, and comprises a service company operating range, a question-answer type and the like. For example, if the target sample customer service is an answer customer service of a small commodity purchase website, the corresponding service environment information is a small commodity purchase background.
Specifically, the keyword set is determined according to the service environment information, that is, keywords which can appear in the interactive question-answering process are determined according to the question-answering background described in the service environment information. The keyword set is a word set capable of reflecting the question-answer content after information simplification is carried out on the question-answer content. And identifying the interactive dialogue group in the customer service interactive corpus by taking the keyword set as an identification index, namely extracting phrases from dialogue texts contained in the interactive dialogue group in the customer service interactive corpus by utilizing the keyword set to obtain the keyword sentence vector. The keyword sentence vector is a vector for carrying out feature description on keywords in the interactive dialogue group.
In one possible embodiment, the keyword set is taken as a screening object, and corpus screening is performed on the interactive dialogue group in the customer service interactive corpus. That is, word matching is performed on the interactive dialogue group text in the customer service interactive corpus by using keywords in the keyword set, and the text which is successfully matched is used as a screening corpus. And further, performing word vector processing on the text in the screening corpus by using a word vector processing module to obtain a vectorization result. Preferably, the texts in the screening corpus are expressed as vectors, so that the similarity relationship between the texts can be obtained according to the similarity relationship between the vectors, and a reliable basis is provided for optimizing the subsequent reply content.
In one embodiment, the word vector of the keyword set and the number of keywords in the interactive dialogue group are utilized, and the sentence vector is calculated by utilizing the keyword sentence vector calculation formula, so that the sentence vector is obtained. And carrying out quantization calculation on the sentence vector by using the keyword sentence vector calculation formula. And obtaining the maximum distance and the minimum distance of the keyword vectors in the two rounds of dialogue according to the obtained keyword sentence vectors, thereby obtaining the first optimal distance interval.
Step S600: and inputting the first optimal distance interval into the NLP recognition module, and optimizing the feedback dialogue content of the target sample customer service.
Further, as shown in fig. 2, the interactive optimization is performed on the target sample customer service according to the first span, and step S600 in this embodiment of the present application further includes:
step S610: acquiring real-time dialogue information of the target sample customer service;
step S620: performing dialogue scene recognition according to the real-time dialogue information, and outputting a first semantic scene, wherein the first semantic scene is a real-time dialogue scene of target sample customer service;
step S630: inputting the first semantic scene into the NLP recognition module for recognition, and outputting a first reply dialogue, wherein the reply dialogue is a dialogue sent by a customer service object to a user object;
step S640: and optimizing the first reply dialogue with the first optimal distance interval, and outputting a second reply dialogue.
Further, step S600 in the embodiment of the present application further includes:
step S650: acquiring a user feedback dialogue of the user object based on the first reply dialogue;
step S660: based on the NLP recognition module, carrying out semantic recognition on the user feedback dialogue, judging whether to activate the interaction optimization instruction according to a semantic recognition result, and if so, carrying out word vector distance positioning by the first distance interval to obtain a second semantic scene;
step S670: and optimizing the first reply dialogue according to the second semantic scene, and outputting the second reply dialogue.
Specifically, data extraction is carried out on the current dialogue of the target sample customer service, and the real-time dialogue information is obtained according to the data extraction result. The real-time dialogue information is data information obtained after the process of carrying out question-answer reply on the target sample customer service is recorded in a text form. By identifying dialogue scenes according to the keywords in the real-time dialogue information, for example, when the keywords in the real-time dialogue information are returned, the first semantic scene is returned. The first semantic scene is a real-time dialogue scene for the target sample customer service, and reflects background information of communication content between a user and the target sample customer service.
In one embodiment, a first reply dialog is obtained by inputting the first semantic scene into the NLP recognition module for natural language processing. The first reply dialogue is the content of intelligent reply to the problem consulted by the user based on the NLP technology, that is, the reply dialogue is the dialogue sent by the customer service object to the user object.
In a possible embodiment, the user object is extracted from the first reply session, and feedback information of the user object on the first reply session, that is, the user feedback session, is obtained. The user feedback dialogue reflects the response of the user object to the first reply dialogue content, and further, semantic recognition is carried out on the user feedback dialogue by using the NLP recognition module, namely intelligent recognition is carried out on information contained in the user feedback dialogue, and whether the interaction optimization instruction needs to be activated is judged according to the information in the semantic recognition result.
In the embodiment of the present application, whether to activate the interaction optimization instruction is determined according to whether the semantic recognition result includes unclear content in the first reply session, if so, it indicates that the user object cannot understand the information included in the first reply session, or the information in the first reply session cannot solve the problem of the user object, and optimization needs to be performed on the first reply session. And according to the interaction optimization instruction, the first optimal distance interval is used as a word vector positioning range, so that word vector distance positioning is performed, and the second semantic scene is obtained according to a positioning result. Wherein the second semantic scene is question-answer background information closer to the question to be consulted by the user object. And optimizing the first reply dialogue based on the NLP technology according to the second semantic scene, so as to obtain a second reply dialogue. The technical effects of improving intelligent customer service interaction quality and shortening feedback time are achieved.
In summary, the embodiments of the present application have at least the following technical effects:
according to the method and the device, the interactive dialogue data set of the target sample customer service is subjected to single-round interactive dialogue identification, so that an interactive dialogue group is obtained, the target of dimension reduction processing on the data set is realized, the processed interactive dialogue group is stored, when interaction optimization is needed, the target of reliable storage on the data is realized, then, the text in the interactive dialogue group is subjected to feature analysis, the interval feature of the keyword vector is obtained, the interval feature is used for obtaining a first optimal distance interval, and the first optimal distance interval is used as the basis for optimizing feedback dialogue of the target sample customer service. The technical effect of improving the accuracy of intelligent customer service interaction and improving the service quality is achieved.
Example two
Based on the same inventive concept as the intelligent customer service interaction method based on NLP in the foregoing embodiments, as shown in FIG. 4, the present application provides an intelligent customer service interaction system based on NLP, and the system and method embodiments in the embodiments of the present application are based on the same inventive concept. Wherein the system comprises:
the data set acquisition module 11 is used for acquiring an interactive dialogue data set of the target sample customer service;
the dialogue group output module 12 is configured to perform single-round interactive dialogue recognition on the interactive data in the interactive dialogue data set by using the interactive objects as segmentation points, and output an interactive dialogue group, where the interactive objects in each group of interactive dialogues at least include a user object and a customer service object;
a dialogue group calling module 13, wherein the dialogue group calling module 13 is used for storing the interactive dialogue group into a customer service interaction corpus, and calling the interactive dialogue group in the customer service interaction corpus when the intelligent customer service interaction system receives an interaction optimization instruction;
the interval feature output module 14 is configured to output interval features that identify keyword vectors in two rounds of conversations by performing feature analysis on the interactive conversation group in the customer service interactive corpus;
the distance interval output module 15 is used for performing keyword vector distance optimization according to the interval characteristics, and outputting a first distance interval, wherein the first distance interval is a maximum distance interval of keyword vectors in two-round conversations;
and the dialogue content optimizing module 16 is configured to input the first distance interval into an NLP identifying module, and optimize the feedback dialogue content of the target sample customer service.
Further, the system further comprises:
the dialogue information acquisition unit is used for acquiring real-time dialogue information of the target sample customer service;
the semantic scene output unit is used for carrying out dialogue scene recognition according to the real-time dialogue information and outputting a first semantic scene, wherein the first semantic scene is a real-time dialogue scene of target sample customer service;
the first reply dialogue output unit is used for inputting the first semantic scene into the NLP recognition module for recognition and outputting a first reply dialogue, wherein the reply dialogue is a dialogue sent by a customer service object to a user object;
and the second reply dialogue output unit is used for optimizing the first reply dialogue in the first range section and outputting a second reply dialogue.
Further, the system further comprises:
a feedback dialogue acquisition unit, configured to acquire a user feedback dialogue of the user object based on the first reply dialogue;
the second semantic scene obtaining unit is used for carrying out semantic recognition on the user feedback dialogue based on the NLP recognition module, judging whether to activate the interaction optimization instruction according to a semantic recognition result, and carrying out word vector distance positioning by the first optimal distance interval if the interaction optimization instruction is activated to obtain a second semantic scene;
and the second dialogue output unit is used for optimizing the first reply dialogue according to the second semantic scene and outputting the second reply dialogue.
Further, the system further comprises:
the interval feature recognition unit is used for carrying out keyword vector interval feature recognition on the target sample customer service by taking a single-round dialogue in the interactive dialogue group as an optimizing unit and taking the interactive dialogue group called in the customer service interactive corpus to output interval period features and interval frequency features;
and the interval characteristic setting unit is used for outputting the interval periodic characteristic and the interval frequency characteristic as the interval characteristic.
Further, the system further comprises:
the service environment information obtaining unit is used for obtaining the service environment information of the target sample customer service;
the keyword set determining unit is used for determining a keyword set according to the service environment information;
the sentence vector output unit is used for identifying the interactive dialogue group in the customer service interactive corpus based on the keyword set and outputting keyword sentence vectors based on the keyword set;
the first distance interval output unit is used for carrying out corpus positioning analysis according to the keyword sentence vectors and outputting a first distance interval.
Further, the system further comprises:
the sentence vector calculation formula setting unit is used for setting a calculation formula of the keyword sentence vector, and the calculation formula is as follows:
wherein ,
characterizing sentence vectors; m represents the number of keywords in the called interactive dialog group; />
A word vector characterizing each keyword; />
The weight of each keyword is characterized.
Further, the system further comprises:
the screening corpus output unit is used for calling the interactive dialogue group in the customer service interactive corpus and outputting a screening corpus based on the keyword set;
and the vectorization result output unit is used for carrying out vectorization representation of words on the screening corpus and outputting vectorization results, wherein the vectorization results comprise word vectors based on the keyword set.
It should be noted that the sequence of the embodiments of the present application is merely for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures 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.
The foregoing description of the preferred embodiments of the present application is not intended to limit the invention to the particular embodiments of the present application, but to limit the scope of the invention to the particular embodiments of the present application.
The specification and drawings are merely exemplary of the application and are to be regarded as covering any and all modifications, variations, combinations, or equivalents that are within the scope of the application. It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the present application and the equivalents thereof, the present application is intended to cover such modifications and variations.
Claims (8)
1. An intelligent customer service interaction method based on NLP is characterized in that the method is applied to an intelligent customer service interaction system, the system comprises an NLP identification module and a word vector processing module, and the method comprises the following steps:
collecting an interactive dialogue data set of a target sample customer service;
carrying out single-round interactive dialogue identification on the interactive data in the interactive dialogue data set by taking the interactive objects as segmentation points, and outputting interactive dialogue groups, wherein the interactive objects in each group of interactive dialogue at least comprise one user object and one customer service object;
storing the interactive dialogue group into a customer service interactive corpus, and calling the interactive dialogue group in the customer service interactive corpus when the intelligent customer service interactive system receives an interactive optimization instruction;
through carrying out feature analysis on the interactive dialogue groups in the customer service interactive corpus, outputting interval features for identifying keyword vectors in two rounds of dialogue;
performing keyword vector distance optimization according to the interval features, and outputting a first optimal distance interval, wherein the first optimal distance interval is a maximum distance interval of keyword vectors in two-round conversations;
and inputting the first optimal distance interval into the NLP recognition module, and optimizing the feedback dialogue content of the target sample customer service.
2. The method of claim 1, wherein the target sample customer service is interactively optimized according to the first span, the method further comprising:
acquiring real-time dialogue information of the target sample customer service;
performing dialogue scene recognition according to the real-time dialogue information, and outputting a first semantic scene, wherein the first semantic scene is a real-time dialogue scene of target sample customer service;
inputting the first semantic scene into the NLP recognition module for recognition, and outputting a first reply dialogue, wherein the reply dialogue is a dialogue sent by a customer service object to a user object;
and optimizing the first reply dialogue with the first optimal distance interval, and outputting a second reply dialogue.
3. The method of claim 2, wherein the method further comprises:
acquiring a user feedback dialogue of the user object based on the first reply dialogue;
based on the NLP recognition module, carrying out semantic recognition on the user feedback dialogue, judging whether to activate the interaction optimization instruction according to a semantic recognition result, and if so, carrying out word vector distance positioning by the first distance interval to obtain a second semantic scene;
and optimizing the first reply dialogue according to the second semantic scene, and outputting the second reply dialogue.
4. A method as claimed in claim 3, wherein the method further comprises:
taking a single-round dialogue in the interactive dialogue group as an optimizing unit, carrying out keyword vector interval feature recognition on the target sample customer service by using the interactive dialogue group called in the customer service interactive corpus, and outputting interval period features and interval frequency features;
and outputting the interval periodic characteristic and the interval frequency characteristic as the interval characteristic.
5. The method of claim 1, wherein the method further comprises:
acquiring service environment information of the target sample customer service;
determining a keyword set according to the service environment information;
identifying the interactive dialogue group in the customer service interactive corpus based on the keyword set, and outputting keyword sentence vectors based on the keyword set;
and carrying out corpus positioning analysis according to the keyword sentence vector, and outputting a first distance interval.
6. The method of claim 5, wherein corpus-location analysis is performed according to the keyword sentence vectors, and a first distance interval is output, and the keyword sentence vectors have the following calculation formula:
wherein ,
characterizing sentence vectors; m represents the number of keywords in the called interactive dialog group; />
A word vector characterizing each keyword; />
The weight of each keyword is characterized.
7. The method of claim 5, wherein the method further comprises:
invoking an interactive dialogue group in the customer service interactive corpus, and outputting a screening corpus based on the keyword set;
and carrying out word vectorization representation on the screening corpus, and outputting vectorization results, wherein the vectorization results comprise word vectors based on the keyword set.
8. An intelligent customer service interaction system based on NLP, the system comprising:
the data set acquisition module is used for acquiring an interactive dialogue data set of the target sample customer service;
the dialogue group output module is used for carrying out single-round interaction dialogue identification on the interaction data in the interaction dialogue data set by taking the interaction objects as the segmentation points and outputting interaction dialogue groups, and the interaction objects in each group of interaction dialogue at least comprise a user object and a customer service object;
the dialogue group calling module is used for storing the interactive dialogue group into a customer service interaction corpus, and calling the interactive dialogue group in the customer service interaction corpus when the intelligent customer service interaction system receives an interaction optimization instruction;
the interval feature output module is used for outputting interval features for identifying keyword vectors in two rounds of conversations by carrying out feature analysis on the interactive conversation groups in the customer service interactive corpus;
the first distance interval is a maximum distance interval of keyword vectors in two-wheel conversations;
and the dialogue content optimizing module is used for inputting the first distance interval into the NLP identifying module and optimizing the feedback dialogue content of the target sample customer service.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310507094.0A CN116226356B (en) | 2023-05-08 | 2023-05-08 | NLP-based intelligent customer service interaction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310507094.0A CN116226356B (en) | 2023-05-08 | 2023-05-08 | NLP-based intelligent customer service interaction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116226356A true CN116226356A (en) | 2023-06-06 |
| CN116226356B CN116226356B (en) | 2023-07-04 |
Family
ID=86571628
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310507094.0A Active CN116226356B (en) | 2023-05-08 | 2023-05-08 | NLP-based intelligent customer service interaction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116226356B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105989550A (en) * | 2015-03-05 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Online service evaluation information determination method and equipment |
| US20180285348A1 (en) * | 2016-07-19 | 2018-10-04 | Tencent Technology (Shenzhen) Company Limited | Dialog generation method, apparatus, and device, and storage medium |
| CN109783623A (en) * | 2018-12-25 | 2019-05-21 | 华东师范大学 | The data analysing method of user and customer service dialogue under a kind of real scene |
| CN111581958A (en) * | 2020-05-27 | 2020-08-25 | 腾讯科技（深圳）有限公司 | Conversation state determining method and device, computer equipment and storage medium |
| CN111694941A (en) * | 2020-05-22 | 2020-09-22 | 腾讯科技（深圳）有限公司 | Reply information determining method and device, storage medium and electronic equipment |
| CN113064980A (en) * | 2021-03-22 | 2021-07-02 | 苏宁金融科技(南京)有限公司 | Intelligent question and answer method and device, computer equipment and storage medium |
| CN113360625A (en) * | 2021-07-02 | 2021-09-07 | 北京容联七陌科技有限公司 | Intelligent dialogue marketing customer acquisition method and system based on NLP |
| WO2022083114A1 (en) * | 2020-10-23 | 2022-04-28 | 中移(上海)信息通信科技有限公司 | Smart dialog method, apparatus, device, storage medium, and program |
| CN114818665A (en) * | 2022-04-22 | 2022-07-29 | 电子科技大学 | Multi-intention identification method and system based on bert + bilstm + crf and xgboost models |
| CN115878768A (en) * | 2022-12-08 | 2023-03-31 | 中国平安财产保险股份有限公司 | NLP-based vehicle insurance service call-back clue recommendation method and related equipment thereof |
-
2023
- 2023-05-08 CN CN202310507094.0A patent/CN116226356B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105989550A (en) * | 2015-03-05 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Online service evaluation information determination method and equipment |
| US20180285348A1 (en) * | 2016-07-19 | 2018-10-04 | Tencent Technology (Shenzhen) Company Limited | Dialog generation method, apparatus, and device, and storage medium |
| CN109783623A (en) * | 2018-12-25 | 2019-05-21 | 华东师范大学 | The data analysing method of user and customer service dialogue under a kind of real scene |
| CN111694941A (en) * | 2020-05-22 | 2020-09-22 | 腾讯科技（深圳）有限公司 | Reply information determining method and device, storage medium and electronic equipment |
| CN111581958A (en) * | 2020-05-27 | 2020-08-25 | 腾讯科技（深圳）有限公司 | Conversation state determining method and device, computer equipment and storage medium |
| WO2022083114A1 (en) * | 2020-10-23 | 2022-04-28 | 中移(上海)信息通信科技有限公司 | Smart dialog method, apparatus, device, storage medium, and program |
| CN113064980A (en) * | 2021-03-22 | 2021-07-02 | 苏宁金融科技(南京)有限公司 | Intelligent question and answer method and device, computer equipment and storage medium |
| CN113360625A (en) * | 2021-07-02 | 2021-09-07 | 北京容联七陌科技有限公司 | Intelligent dialogue marketing customer acquisition method and system based on NLP |
| CN114818665A (en) * | 2022-04-22 | 2022-07-29 | 电子科技大学 | Multi-intention identification method and system based on bert + bilstm + crf and xgboost models |
| CN115878768A (en) * | 2022-12-08 | 2023-03-31 | 中国平安财产保险股份有限公司 | NLP-based vehicle insurance service call-back clue recommendation method and related equipment thereof |
Non-Patent Citations (1)
| Title |
|---|
| 李伟通;皮德常;: "基于统计学习的自然语言对话系统的设计与实现", 微计算机应用, no. 07 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116226356B (en) | 2023-07-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117033608B (en) | Knowledge graph generation type question-answering method and system based on large language model | |
| CN110046221B (en) | Machine dialogue method, device, computer equipment and storage medium | |
| CN110188358B (en) | Training method and device for natural language processing model | |
| CN111916111B (en) | Intelligent voice outbound method and device with emotion, server and storage medium | |
| CN112100354B (en) | Man-machine conversation method, device, equipment and storage medium | |
| US20160210279A1 (en) | Methods and systems for analyzing communication situation based on emotion information | |
| CN110019742B (en) | Method and device for processing information | |
| CN113268609B (en) | Knowledge graph-based dialogue content recommendation method, device, equipment and medium | |
| KR102100214B1 (en) | Method and appratus for analysing sales conversation based on voice recognition | |
| CN112885336B (en) | Training and recognition method and device of voice recognition system and electronic equipment | |
| CN111339302A (en) | Method and device for training element classification model | |
| CN111813909A (en) | Intelligent question answering method and device | |
| CN111858854A (en) | Question-answer matching method based on historical dialogue information and related device | |
| CN114678014A (en) | Intention recognition method, device, computer equipment and computer readable storage medium | |
| CN114218945A (en) | Entity identification method, device, server and storage medium | |
| CN113342958A (en) | Question-answer matching method, text matching model training method and related equipment | |
| CN115497465A (en) | Voice interaction method and device, electronic equipment and storage medium | |
| CN110717027B (en) | Multi-round intelligent question-answering method, system, controller and medium | |
| CN116361442B (en) | Business hall data analysis method and system based on artificial intelligence | |
| CN110795531B (en) | Intention identification method, device and storage medium | |
| CN113051384A (en) | User portrait extraction method based on conversation and related device | |
| CN115186071A (en) | Intention recognition method and device, electronic equipment and readable storage medium | |
| CN116226356B (en) | NLP-based intelligent customer service interaction method and system | |
| CN114238605B (en) | Automatic conversation method and device for intelligent voice customer service robot | |
| CN113111855B (en) | Multi-mode emotion recognition method and device, electronic equipment 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 |