CN112988948A - Service processing method and device - Google Patents
Service processing method and device Download PDFInfo
- Publication number
- CN112988948A CN112988948A CN202110160323.7A CN202110160323A CN112988948A CN 112988948 A CN112988948 A CN 112988948A CN 202110160323 A CN202110160323 A CN 202110160323A CN 112988948 A CN112988948 A CN 112988948A
- Authority
- CN
- China
- Prior art keywords
- text
- target
- information
- service
- question
- 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
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
-
- 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
-
- 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
Abstract
An embodiment of the present specification provides a service processing method and a device, where the service processing method includes: obtaining historical multimedia service data containing at least two conversation roles in a target service, converting the historical multimedia service data into text information, dividing the text information into text segments, dividing the text segments into conversation roles according to semantic information of the text segments, determining the target conversation roles according to a conversation role division result and the semantic information, obtaining an incidence relation between problem texts of the target conversation roles in the text segments, screening target problem texts of the target conversation roles in the text segments according to the incidence relation, and constructing a conversational knowledge base of the target service based on the target problem texts.
Description
Technical Field
The embodiment of the specification relates to the technical field of computers, in particular to a service processing method. One or more embodiments of the present specification also relate to a business processing apparatus, a computing device, and a computer-readable storage medium.
Background
Along with the increase of the attention of user groups to health care projects, the number of people who join the projects is more and more, so that in order to enable users to experience or enjoy the services provided by the projects more quickly and conveniently, different from a traditional offline service mode, a plurality of online service platforms are produced accordingly. The user can join the health care project through the offline or online service platform, and under the condition that the project initiator needs to provide health care for the user when the user joins the project and health problems occur, the project initiator of the health care project needs to collect the clinic evidence and other related proving materials of the user, and the clinic evidence and other proving materials are audited to judge whether the user meets the condition of enjoying the health care.
At present, most service initiators adopt a pre-constructed knowledge base to generate a collection certification material, but at present, the construction of the knowledge base completely depends on manual work or manual work assisted by mining high-frequency customer messages, the initial labor cost is very high, and since knowledge needs to be updated iteratively, the alternation of the knowledge base also needs to be maintained continuously by manual work, which results in high maintenance cost, and therefore, an effective method is urgently needed to be provided to solve the problems.
Disclosure of Invention
In view of this, the embodiments of the present specification provide a service processing method. One or more embodiments of the present disclosure also relate to a business processing apparatus, a computing device, and a computer-readable storage medium to solve technical problems in the prior art.
According to a first aspect of embodiments of the present specification, there is provided a service processing method, including:
obtaining historical multimedia service data containing at least two conversation roles in a target service, and converting the historical multimedia service data into text information;
dividing the text information into text segments, carrying out conversation role division on the text segments according to semantic information of the text segments, and determining a target conversation role according to a conversation role division result and the semantic information;
acquiring the incidence relation between the question texts of the target conversation roles in the text segments;
and screening the target problem text of the target dialogue role in the text fragment according to the incidence relation, and constructing a dialect knowledge base of the target service based on the target problem text.
Optionally, the converting the historical multimedia service data into text information includes:
and performing voice recognition on the historical multimedia service data to obtain text information generated by recognition.
Optionally, the dividing the text information into text segments includes:
and determining a semantic matching result of at least one service information acquisition item and the text information in the service processing template of the target service, and dividing the text information into at least one text segment according to the semantic matching result.
Optionally, the determining a semantic matching result between at least one service information acquisition item and the text information in the service processing template of the target service includes:
determining information to be acquired corresponding to at least one service information acquisition item based on at least one service information acquisition item in the service processing template of the target service;
semantic similarity calculation is carried out on the information to be collected and the text information;
and determining the semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
Optionally, the determining a semantic matching result between at least one service information acquisition item and the text information in the service processing template of the target service includes:
determining information to be acquired corresponding to at least one service information acquisition item based on at least one service information acquisition item in the service processing template of the target service;
inputting the information to be acquired and the text information into a pre-trained semantic matching model for similarity calculation;
and determining the semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
Optionally, the obtaining of the association relationship between the question texts of the target dialog roles in the text segment includes:
inputting the text segments into a natural language processing model for relevance calculation to generate relevance calculation results among question texts of the target conversation roles in the text segments;
and determining the association relation between the question texts of the target conversation roles in the text segments according to the association degree calculation result.
Optionally, the obtaining of the association relationship between the question texts of the target dialog roles in the text segment includes:
acquiring the ith question text of the target dialog role, the answer text of the ith question text and the semantic information of the (i + 1) th question text in the text fragment;
and determining the incidence relation between the ith question text and the (i + 1) th question text based on the semantic information, wherein i belongs to [1, n-1], n is the number of question texts of a target conversation role in the text segment, and n is a positive integer.
Optionally, the screening, according to the association relationship, a target question text of the target dialog role in the text fragment includes:
determining the text length of the problem text with the association relation and/or the text number of the problem text with the association relation;
and screening the target question text of the target conversation role according to the text length and/or the text quantity.
Optionally, the screening, according to the association relationship, a target question text of the target dialog role in the text fragment includes:
forming a text pair by the question text with the incidence relation and the answer text corresponding to the question text, and calculating the information entropy of the text pair;
and screening the target problem text of the target conversation role in the text segment according to the information entropy.
Optionally, the service processing method further includes:
collecting voice data of a user, and performing semantic analysis on the voice data;
and screening the dialect recommendation information matched with the voice data in the dialect knowledge base according to a semantic analysis result, and recommending the dialect recommendation information to the user.
According to a second aspect of embodiments herein, there is provided a service processing apparatus, including:
the system comprises an acquisition module, a processing module and a display module, wherein the acquisition module is configured to acquire historical multimedia service data containing at least two conversation roles in a target service and convert the historical multimedia service data into text information;
the dividing module is configured to divide the text information into text segments, perform conversation role division on the text segments according to semantic information of the text segments, and determine a target conversation role according to a conversation role division result and the semantic information;
the acquisition module is configured to acquire the incidence relation among the question texts of the target conversation roles in the text segments;
and the construction module is configured to screen the target question texts of the target conversation roles in the text segments according to the incidence relation and construct a conversational knowledge base of the target service based on the target question texts.
According to a third aspect of embodiments herein, there is provided a computing device comprising:
a memory and a processor;
the memory is configured to store computer-executable instructions and the processor is configured to execute the computer-executable instructions to implement the steps of the business process method.
According to a fourth aspect of embodiments herein, there is provided a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the business processing method.
One embodiment of the present specification obtains historical multimedia service data including at least two conversation roles in a target service, converts the historical multimedia service data into text information, divides the text information into text segments, performs conversation role division on the text segments according to semantic information of the text segments, determines a target conversation role according to a conversation role division result and the semantic information, obtains an association relationship between problem texts of the target conversation role in the text segments, screens a target problem text of the target conversation role in the text segments according to the association relationship, and constructs a conversational knowledge base of the target service based on the target problem text.
The dialect knowledge base is constructed through the historical multimedia service data of the target service, so that the construction complexity of the dialect knowledge base and the maintenance difficulty of the knowledge base are reduced, and the construction efficiency is improved; in addition, text segments of the text information of the historical multimedia service data are divided, and the dialect knowledge base of the target service is constructed according to different text segments, so that the coverage rate of the dialect knowledge base is improved, and the accuracy of the dialect contained in the constructed dialect knowledge base is further ensured.
Drawings
Fig. 1 is a processing flow diagram of a service processing method provided in an embodiment of the present specification;
fig. 2 is a schematic diagram of a service processing procedure provided in an embodiment of the present specification;
fig. 3 is a flowchart illustrating a processing procedure of a service processing method according to an embodiment of the present disclosure;
fig. 4 is a schematic diagram of a service processing apparatus provided in an embodiment of the present specification;
fig. 5 is a block diagram of a computing device according to an embodiment of the present disclosure.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present description. This description may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein, as those skilled in the art will be able to make and use the present disclosure without departing from the spirit and scope of the present disclosure.
The terminology used in the description of the one or more embodiments is for the purpose of describing the particular embodiments only and is not intended to be limiting of the description of the one or more embodiments. As used in one or more embodiments of the present specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used in one or more embodiments of the present specification refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It will be understood that, although the terms first, second, etc. may be used herein in one or more embodiments to describe various information, these information should not be limited by these terms. These terms are only used to distinguish one type of information from another. For example, a first can also be referred to as a second and, similarly, a second can also be referred to as a first without departing from the scope of one or more embodiments of the present description. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.
In the present specification, a service processing method is provided, and the present specification relates to a service processing apparatus, a computing device, and a computer-readable storage medium, which are described in detail in the following embodiments one by one.
Fig. 1 shows a processing flow chart of a service processing method provided according to an embodiment of the present specification, which includes steps 102 to 108.
Specifically, the historical multimedia service data includes, but is not limited to, video data or audio data, and the target service refers to a service to be audited, and includes, but is not limited to, a claim service to be audited, a loan service to be audited, an investment service to be audited, and the like.
If the target service is a service to be claim settled, after the user participates in insurance and goes out of risk, a claim settlement application is sent to the service initiator, the service initiator collects the medical examination certificates and other related certification materials of the user through interview or investigation and the like, and the medical examination certificates and other certification materials are initially checked to prepare for subsequently judging whether the user meets the claim settlement conditions. The historical multimedia service data is video data or audio data formed by recording the interview process in a video or audio mode in the process of collecting related materials by the service initiator in an interview mode, wherein the video data or the audio data comprise at least two conversation roles, and the at least two conversation roles can comprise the service initiator, the user and the like.
In specific implementation, after the historical multimedia service data is obtained, the historical multimedia service data needs to be converted into text information, a jargon knowledge base is built based on the text information, the historical multimedia service data is converted into the text information, specifically, voice recognition is performed on the historical multimedia service data, and the text information generated by recognition is obtained.
In practical applications, voice information in the historical multimedia service data may be converted into text information using an Automatic Speech Recognition (ASR) technology.
And 104, dividing the text information into text segments, dividing the text segments into conversation roles according to semantic information of the text segments, and determining a target conversation role according to a conversation role division result and the semantic information.
Specifically, the target service is a service to be audited, in order to ensure accuracy of an audit result in the process of auditing the service to be audited, multiple items of service information are usually audited, a final audit result is obtained comprehensively, and relevance between any two or more items of service information in the multiple items of service information which may be audited is low, so that if the text information is taken as a whole to perform relevance analysis of a problem text, accuracy of an obtained analysis result is not high, and therefore, in the process of constructing the technical knowledge base, the text information can be divided according to matching relations between service information acquisition items and the text information in a service processing template, so that the technical knowledge base corresponding to each service information acquisition item is constructed according to each text segment generated by division.
As mentioned above, the historical multimedia service data is video data or audio data formed by recording the interview process in a video or audio mode, and the video data or audio data includes at least two conversation roles, namely a service initiator and a user.
The embodiment of the specification aims to construct a dialect knowledge base for a target service according to historical multimedia service data, and recommend a dialect link for a service initiator by using the dialect knowledge base, so that more effective auditing material information is mined through the dialect information in the dialect link.
In specific implementation, the text information is divided into text segments, that is, semantic matching results of at least one service information acquisition item and the text information in a service processing template of the target service are determined, and the text information is divided into at least one text segment according to the semantic matching results.
Further, determining a semantic matching result between at least one service information acquisition item and the text information in the service processing template of the target service can be specifically realized by the following steps:
determining information to be acquired corresponding to at least one service information acquisition item based on at least one service information acquisition item in the service processing template of the target service;
semantic similarity calculation is carried out on the information to be collected and the text information;
and determining the semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
Specifically, since the target service is a service to be audited, in the process of acquiring audit data by a face visit, a service initiator of the target service can acquire the audit data by asking a user based on a service processing template, and since acquisition information included in historical multimedia service data obtained by recording the face visit process generally corresponds to each information acquisition item in the service processing template, text segment division can be performed on text information by determining semantic matching results of each service information acquisition item in the service processing template and the text information according to the semantic matching results. And then, a dialect knowledge base corresponding to each information acquisition item can be constructed through the text information in each text segment.
Still taking the target service as the claim service to be checked as an example, in the interview process, the service initiator will ask a question to the user based on the service processing template (the claim information collecting template of the claim service), and collect the claim material based on the reply information of the user. Since the service information collection items in the information collection template generally include user basic information (name, age, etc.), a business insurance, a social security, a physical examination, hypertension/diabetes/hepatitis, etc., the text information can be segmented by determining semantic matching results of each service information collection item and the text information and according to the semantic matching results.
In practical application, the information to be acquired corresponding to each service information acquisition item can be determined based on the service information acquisition items in the service processing template, and the semantic similarity between the information to be acquired and the text information is calculated, so that the text information is divided into text segments according to the calculation result of the semantic similarity.
For example, if the converted text information includes 10 lines of data, and it is determined that the 1 st to 4 th lines of text are related to the basic information of the user, the 5 th to 7 th lines are related to social security, and the 8 th to 10 th lines are related to physical examination according to the semantic similarity calculation result, the text information is divided into 3 text segments (the 1 st to 4 lines are divided into one text segment, the 5 th to 7 lines are divided into one text segment, and the 8 th to 10 lines are divided into one text segment).
In the process of constructing the jargon knowledge base, the text information is divided according to the matching relation between the business information acquisition items and the text information in the business processing template, so that the jargon knowledge base corresponding to each business information acquisition item is respectively constructed according to each text segment generated by division, and the effectiveness and the accuracy of the constructed jargon knowledge base of each business information acquisition item are favorably ensured.
In addition, determining the semantic matching result between at least one service information acquisition item and the text information in the service processing template of the target service can be realized by the following method:
determining information to be acquired corresponding to at least one service information acquisition item based on at least one service information acquisition item in the service processing template of the target service;
inputting the information to be acquired and the text information into a pre-trained semantic matching model for similarity calculation;
and determining the semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
Specifically, the similarity between the information to be collected and the text information may be calculated by using an Enhanced Sequential Inference Model (ESIM).
The embodiment of the specification performs similarity calculation on the information to be acquired and the text information by means of the semantic matching model, so that the accuracy of a calculation result is ensured, and the efficiency of a calculation process is improved.
In addition, in practical application, in the interview process, the service initiator collects the audit information of the user mostly by asking the user, so that the target conversation role can be determined by determining the text sentence pattern, and specifically, the conversation role with more question sentence patterns in the text of each conversation role can be determined as the target conversation role.
And 106, acquiring the association relation between the question texts of the target conversation roles in the text segments.
Specifically, in the embodiment of the present specification, the conversational knowledge base of the target service needs to be constructed through the historical multimedia service data of the target service, so that in the actual application process, a conversational link that can be used in the interview process is recommended to the user according to the association relationship between the conversational knowledge bases, and thus, more effective audit material information is mined through conversational information in the conversational link.
Therefore, in the process of constructing the dialogistic knowledge base, the association relationship between the question texts of the target dialog roles in the text segments needs to be constructed, so as to generate the dialogistic link according to the association relationship.
In specific implementation, the association relationship between the question texts of the target dialog role in the text segment is obtained, specifically, the text segment can be input into a natural language processing model for association calculation, a calculation result of the association degree between the question texts of the target dialog role in the text segment is generated, and the association relationship between the question texts of the target dialog role in the text segment is determined according to the calculation result of the association degree.
Specifically, the relevance of the question texts in the text segment can be calculated through a Natural Language Processing (NLP) model, and the relevance calculation result between the question texts of the target dialog role in the text segment is generated, so that the relevance between the question texts is determined according to the relevance calculation result.
In addition, the incidence relation between all question texts of the target dialog role in the whole text segment can be determined by determining the incidence relation between any two adjacent question texts in the text segment, and the method can be specifically realized by the following steps:
acquiring the ith question text of the target dialog role, the answer text of the ith question text and the semantic information of the (i + 1) th question text in the text fragment;
and determining the incidence relation between the ith question text and the (i + 1) th question text based on the semantic information, wherein i belongs to [1, n-1], n is the number of question texts of a target conversation role in the text segment, and n is a positive integer.
Specifically, the text segment includes n question texts of the target dialog role, and the incidence relation between the question texts of the target dialog role in the whole text segment can be determined by determining whether the incidence relation exists between any two adjacent question texts.
For example, if the text fragment includes 5 question texts of the target dialog character, wherein if it is determined whether an association relationship exists between the 2 nd question text and the 3 rd question text of the target dialog character, the association relationship is specifically determined according to the 2 nd question text, the 3 rd question text, and an answer text corresponding to the 2 nd question text of the target dialog character.
In the actual interview process, in order to ensure the validity of the audit information obtained by interview, the service initiator needs to adjust the next interview question information in real time according to the response information of the user, and a specific adjustment strategy needs to be given by the conversational knowledge base, so that in the process of constructing the conversational knowledge base, the incidence relation among the question texts of the target conversation role in the text segment needs to be determined, and a foundation is laid for generating a conversational link which can be referred by the service initiator according to the incidence relation.
And 108, screening the target problem text of the target conversation role in the text segment according to the incidence relation, and constructing a conversational knowledge base of the target service based on the target problem text.
Specifically, after the incidence relation between the question texts of the target conversation role in the text segment is determined, the target texts can be further screened according to the incidence relation to construct a conversational knowledge base.
In specific implementation, the target question text of the target dialog role in the text fragment is screened according to the association relationship, and the method can be specifically realized by the following steps:
determining the text length of the problem text with the association relation and/or the text number of the problem text with the association relation;
and screening the target question text of the target conversation role according to the text length and/or the text quantity.
In addition, the target question text of the target dialog role in the text fragment is screened according to the incidence relation, and the method can also be realized by the following mode:
forming a text pair by the question text with the incidence relation and the answer text corresponding to the question text, and calculating the information entropy of the text pair;
and screening the target problem text of the target conversation role in the text segment according to the information entropy.
Specifically, the quality of the target question text can be evaluated from three dimensions of text length, number of question texts and information entropy;
the length of the answer can be intuitively reflected by the length of the answer, and in general, the longer the answer of a possible user is, the more effective information the user may contain, so that the target question text can be jointly screened by combining the text lengths of the question text and the answer text corresponding to the question text, and particularly, the question text in which the sum of the text lengths of the question text and the answer text is greater than a preset length threshold value in the question text with the association relationship can be used as the target question text.
In addition, the number of the question texts can be used for reflecting the number of question-answer interactions between the service initiator and the user, and the larger the number of the question texts is, the larger the number of the question-answer interactions between the service initiator and the user is, the deeper the depth of the question texts can be further indicated, and the more effective the information contained in the corresponding answer texts is, so that one, two or more groups of question texts of which the number of texts is greater than a preset number threshold value in a plurality of groups of question texts with an association relationship can be used as the target question text.
In addition, since the information entropy is a measure of event uncertainty, if a question text and an answer text form a text pair, the information entropy of the text pair can be used for representing the accuracy of the answer text relative to the question text, and the smaller the information entropy is, the more accurate the answer is indicated, and the more effective the question is indicated, so that the question text in the text pair formed by the question text with an association relationship and the corresponding answer text can be used as the target question text.
Or if the problem texts are screened from the three dimensions of the text length, the text quantity and the information entropy, the weighted calculation can be performed according to the weights corresponding to the three dimensions respectively, so as to screen the target problem texts.
And screening the target problem text from three dimensions of text length, text quantity and information entropy, and facilitating the improvement of the effectiveness of the target problem text obtained by screening.
In addition, after the target problem text is obtained through screening, the target problem text needs to be checked to determine whether the target problem text is matched with the dialect in the initial dialect script constructed manually in advance, and the initial dialect script is updated according to the matching result.
For example, the initial conversational script does not include a node for inquiring whether you have hypertension, but the target question text screened finally by processing the historical multimedia service data includes "whether you have hypertension", and the semantic matching degree of the target question text and the inquiring module about hypertension in the initial conversational script is high, the target question text of "whether you have hypertension" can be added to the inquiring module about hypertension in the initial conversational script, so as to update the initial conversational script.
After the initial conversational script is updated to generate a conversational knowledge base, the conversational knowledge base can be stored to a service platform for a service initiator to use, and the conversational knowledge base can be continuously updated in a later period.
In addition, the method can also be used for carrying out speech recommendation for a service initiator based on a speech technology knowledge base, and by acquiring voice data of a user and carrying out semantic analysis on the voice data, speech technology recommendation information matched with the voice data is screened in the speech technology knowledge base according to a semantic analysis result and recommended to the user.
By adopting the method, the speech technology is recommended for the service initiator, so that the service initiator can recommend the speech technology with higher quality in real time in the interview process of the service initiator, and the service initiator can acquire more effective audit material information.
A schematic diagram of a service processing process provided in an embodiment of the present specification is shown in fig. 2, and the method includes first obtaining historical multimedia service data (audio file) of a target service, converting the audio file into text information (voice to text), and then cutting the text information according to semantic information of each information acquisition item in a service processing template to generate a plurality of text segments; then screening high-quality dialogues from each text segment, grading the high-quality dialogues in the screening result, determining a target dialogues according to the grading result, verifying the target dialogues, and updating an initial dialogues script by using the target dialogues under the condition that the verification is passed (the initial dialogues script is constructed manually based on information acquisition items such as user basic information, business/social security, physical examination, past medical history and the like); and storing the updated dialogs, and performing iterative optimization on the stored dialogs at a later stage by evaluating the stored dialogs so as to construct a closed loop optimized by the generation, storage and evaluation of the dialogs.
The dialect knowledge base is constructed through the historical multimedia service data of the target service, so that the construction complexity of the dialect knowledge base and the maintenance difficulty of the knowledge base are reduced, and the construction efficiency is improved; in addition, text segments of the text information of the historical multimedia service data are divided, and the dialect knowledge base of the target service is constructed according to different text segments, so that the coverage rate of the dialect knowledge base is improved, and the accuracy of the dialect contained in the constructed dialect knowledge base is further ensured.
The following description further describes the service processing method by taking an application of the service processing method provided in this specification in a claim settlement scenario as an example, with reference to fig. 3. Fig. 3 shows a flowchart of a processing procedure of a service processing method according to an embodiment of the present specification, and specific steps include step 302 to step 322.
Step 302, obtaining historical interview video data containing at least two conversation roles in the claim settlement service.
And 304, performing voice recognition on the historical interview video data to obtain text information generated by recognition.
And 308, inputting the information to be acquired and the text information into a pre-trained semantic matching model for similarity calculation.
And 310, determining a semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
Step 312, dividing the text information into at least one text segment according to the semantic matching result.
And 314, carrying out conversation role division on the text fragments according to the semantic information of the text fragments, and determining a target conversation role according to a conversation role division result and the semantic information.
And step 316, inputting the text segments into a natural language processing model for relevance calculation, and generating a relevance calculation result between the question texts of the target conversation roles in the text segments.
Step 318, determining the association relation between the question texts of the target conversation roles in the text segments according to the association degree calculation result.
And 320, screening the target question text of the target conversation role in the text fragment according to the incidence relation.
In the embodiment of the specification, the dialect knowledge base is constructed through the historical interview video data of the claim settlement service, so that the construction complexity of the dialect knowledge base and the maintenance difficulty of the knowledge base are reduced, and the construction efficiency is improved; in addition, text segments of the text information of the historical interview video data are divided, and a dialect knowledge base corresponding to different service information acquisition items is constructed according to different text segments, so that the coverage rate of the dialect knowledge base is improved, and the accuracy of dialects contained in the constructed dialect knowledge base is further ensured.
Corresponding to the above method embodiment, the present specification further provides an embodiment of a service processing apparatus, and fig. 4 shows a schematic diagram of a service processing apparatus provided in an embodiment of the present specification. As shown in fig. 4, the apparatus includes:
a conversion module 402, configured to obtain historical multimedia service data including at least two conversation roles in a target service, and convert the historical multimedia service data into text information;
a dividing module 404 configured to divide the text information into text segments, perform dialog role division on the text segments according to semantic information of the text segments, and determine a target dialog role according to a dialog role division result and the semantic information;
an obtaining module 406, configured to obtain an association relationship between question texts of the target conversation roles in the text segments;
the building module 408 is configured to filter the target question texts of the target conversation roles in the text segments according to the association relationship, and build a conversational knowledge base of the target service based on the target question texts.
Optionally, the conversion module 402 includes:
and the voice recognition submodule is configured to perform voice recognition on the historical multimedia service data to obtain text information generated by recognition.
Optionally, the dividing module 404 includes:
and the dividing submodule is configured to determine a semantic matching result of at least one service information acquisition item and the text information in a service processing template of the target service, and divide the text information into at least one text segment according to the semantic matching result.
Optionally, the partitioning sub-module includes:
the first determining unit is configured to determine information to be acquired corresponding to at least one service information acquisition item based on the at least one service information acquisition item in the service processing template of the target service;
the first calculation unit is configured to perform semantic similarity calculation on the information to be acquired and the text information;
and the second determining unit is configured to determine a semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
Optionally, the partitioning sub-module includes:
the third determining unit is configured to determine information to be acquired corresponding to at least one service information acquisition item based on the at least one service information acquisition item in the service processing template of the target service;
the second calculation unit is configured to perform similarity calculation on the information to be acquired and the pre-trained semantic matching model input by the text information;
and the fourth determining unit is configured to determine a semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
Optionally, the obtaining module 406 includes:
the relevance operator module is configured to input the text segments into a natural language processing model for relevance calculation, and relevance calculation results among the question texts of the target conversation roles in the text segments are generated;
and the first obtaining submodule is configured to determine the association relation between the question texts of the target conversation roles in the text segments according to the association degree calculation result.
Optionally, the obtaining module 406 includes:
the second obtaining submodule is configured to obtain the ith question text of the target conversation role, the answer text of the ith question text and the semantic information of the (i + 1) th question text in the text fragment;
and a third determining submodule configured to determine an association relationship between the ith question text and the (i + 1) th question text based on the semantic information, wherein i belongs to [1, n-1], n is the number of question texts of a target conversation role in the text segment, and n is a positive integer.
Optionally, the building module 408 includes:
the text information determining sub-module is configured to determine the text length of the problem text with the association relation and/or the text number of the problem text with the association relation;
and the first screening submodule is configured to screen the target question text of the target conversation role according to the text length and/or the text quantity.
Optionally, the building module 408 includes:
the information entropy calculation sub-module is configured to combine the question texts with the incidence relations and the answer texts corresponding to the question texts into text pairs and calculate the information entropy of the text pairs;
and the second screening submodule is configured to screen the target question text of the target conversation role in the text segment according to the information entropy.
Optionally, the service processing apparatus further includes:
the acquisition module is configured to acquire voice data of a user and perform semantic analysis on the voice data;
and the recommendation module is configured to screen the dialect recommendation information matched with the voice data in the dialect knowledge base according to a semantic analysis result and recommend the dialect recommendation information to the user.
The foregoing is a schematic scheme of a service processing apparatus according to this embodiment. It should be noted that the technical solution of the service processing apparatus and the technical solution of the service processing method belong to the same concept, and details that are not described in detail in the technical solution of the service processing apparatus can be referred to the description of the technical solution of the service processing method.
FIG. 5 illustrates a block diagram of a computing device 500 provided in accordance with one embodiment of the present description. The components of the computing device 500 include, but are not limited to, a memory 510 and a processor 520. Processor 520 is coupled to memory 510 via bus 530, and database 550 is used to store data.
Computing device 500 also includes access device 540, access device 540 enabling computing device 500 to communicate via one or more networks 560. Examples of such networks include the Public Switched Telephone Network (PSTN), a Local Area Network (LAN), a Wide Area Network (WAN), a Personal Area Network (PAN), or a combination of communication networks such as the internet. The access device 540 may include one or more of any type of network interface, e.g., a Network Interface Card (NIC), wired or wireless, such as an IEEE802.11 Wireless Local Area Network (WLAN) wireless interface, a worldwide interoperability for microwave access (Wi-MAX) interface, an ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a bluetooth interface, a Near Field Communication (NFC) interface, and so forth.
In one embodiment of the present description, the above-described components of computing device 500, as well as other components not shown in FIG. 5, may also be connected to each other, such as by a bus. It should be understood that the block diagram of the computing device architecture shown in FIG. 5 is for purposes of example only and is not limiting as to the scope of the present description. Those skilled in the art may add or replace other components as desired.
Computing device 500 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., tablet, personal digital assistant, laptop, notebook, netbook, etc.), mobile phone (e.g., smartphone), wearable computing device (e.g., smartwatch, smartglasses, etc.), or other type of mobile device, or a stationary computing device such as a desktop computer or PC. Computing device 500 may also be a mobile or stationary server.
Wherein the memory 510 is configured to store computer-executable instructions and the processor 520 is configured to execute the computer-executable instructions for implementing the steps of the business process method.
The above is an illustrative scheme of a computing device of the present embodiment. It should be noted that the technical solution of the computing device and the technical solution of the service processing method belong to the same concept, and details that are not described in detail in the technical solution of the computing device can be referred to the description of the technical solution of the service processing method.
An embodiment of the present specification also provides a computer readable storage medium storing computer instructions, which when executed by a processor, are used for implementing the steps of the service processing method.
The above is an illustrative scheme of a computer-readable storage medium of the present embodiment. It should be noted that the technical solution of the storage medium belongs to the same concept as the technical solution of the service processing method, and details that are not described in detail in the technical solution of the storage medium can be referred to the description of the technical solution of the service processing method.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may 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 may also be possible or may be advantageous.
The computer instructions comprise computer program code which may be in the form of source code, object code, an executable file or some intermediate form, or the like. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like. It should be noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals as is required by legislation and patent practice.
It should be noted that, for the sake of simplicity, the foregoing method embodiments are described as a series of acts, but those skilled in the art should understand that the present embodiment is not limited by the described acts, because some steps may be performed in other sequences or simultaneously according to the present embodiment. Further, those skilled in the art should also appreciate that the embodiments described in this specification are preferred embodiments and that acts and modules referred to are not necessarily required for an embodiment of the specification.
In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
The preferred embodiments of the present specification disclosed above are intended only to aid in the description of the specification. Alternative embodiments are not exhaustive and do not limit the invention to the precise embodiments described. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the embodiments and the practical application, to thereby enable others skilled in the art to best understand and utilize the embodiments. The specification is limited only by the claims and their full scope and equivalents.
Claims (13)
1. A service processing method comprises the following steps:
obtaining historical multimedia service data containing at least two conversation roles in a target service, and converting the historical multimedia service data into text information;
dividing the text information into text segments, carrying out conversation role division on the text segments according to semantic information of the text segments, and determining a target conversation role according to a conversation role division result and the semantic information;
acquiring the incidence relation between the question texts of the target conversation roles in the text segments;
and screening the target problem text of the target dialogue role in the text fragment according to the incidence relation, and constructing a dialect knowledge base of the target service based on the target problem text.
2. The traffic processing method according to claim 1, wherein said converting the historical multimedia traffic data into text information comprises:
and performing voice recognition on the historical multimedia service data to obtain text information generated by recognition.
3. The traffic processing method according to claim 1 or 2, wherein said dividing the text information into text segments comprises:
and determining a semantic matching result of at least one service information acquisition item and the text information in the service processing template of the target service, and dividing the text information into at least one text segment according to the semantic matching result.
4. The business processing method of claim 3, wherein determining the semantic matching result between at least one business information collection item and the text information in the business processing template of the target business comprises:
determining information to be acquired corresponding to at least one service information acquisition item based on at least one service information acquisition item in the service processing template of the target service;
semantic similarity calculation is carried out on the information to be collected and the text information;
and determining the semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
5. The business processing method of claim 3, wherein determining the semantic matching result between at least one business information collection item and the text information in the business processing template of the target business comprises:
determining information to be acquired corresponding to at least one service information acquisition item based on at least one service information acquisition item in the service processing template of the target service;
inputting the information to be acquired and the text information into a pre-trained semantic matching model for similarity calculation;
and determining the semantic matching result of the at least one service information acquisition item and the text information according to the calculation result.
6. The service processing method according to claim 1 or 2, wherein the obtaining of the association relationship between the question texts of the target conversation roles in the text segments includes:
inputting the text segments into a natural language processing model for relevance calculation to generate relevance calculation results among question texts of the target conversation roles in the text segments;
and determining the association relation between the question texts of the target conversation roles in the text segments according to the association degree calculation result.
7. The service processing method according to claim 1 or 2, wherein the obtaining of the association relationship between the question texts of the target conversation roles in the text segments includes:
acquiring the ith question text of the target dialog role, the answer text of the ith question text and the semantic information of the (i + 1) th question text in the text fragment;
and determining the incidence relation between the ith question text and the (i + 1) th question text based on the semantic information, wherein i belongs to [1, n-1], n is the number of question texts of a target conversation role in the text segment, and n is a positive integer.
8. The service processing method according to claim 1, wherein the screening of the target question text of the target dialog role in the text fragment according to the association relationship comprises:
determining the text length of the problem text with the association relation and/or the text number of the problem text with the association relation;
and screening the target question text of the target conversation role according to the text length and/or the text quantity.
9. The service processing method according to claim 1 or 8, wherein the screening of the target question text of the target dialog role in the text fragment according to the association relationship comprises:
forming a text pair by the question text with the incidence relation and the answer text corresponding to the question text, and calculating the information entropy of the text pair;
and screening the target problem text of the target conversation role in the text segment according to the information entropy.
10. The traffic processing method according to claim 1, further comprising:
collecting voice data of a user, and performing semantic analysis on the voice data;
and screening the dialect recommendation information matched with the voice data in the dialect knowledge base according to a semantic analysis result, and recommending the dialect recommendation information to the user.
11. A traffic processing apparatus, comprising:
the conversion module is configured to acquire historical multimedia service data containing at least two conversation roles in a target service and convert the historical multimedia service data into text information;
the dividing module is configured to divide the text information into text segments, perform conversation role division on the text segments according to semantic information of the text segments, and determine a target conversation role according to a conversation role division result and the semantic information;
the acquisition module is configured to acquire the incidence relation among the question texts of the target conversation roles in the text segments;
and the construction module is configured to screen the target question texts of the target conversation roles in the text segments according to the incidence relation and construct a conversational knowledge base of the target service based on the target question texts.
12. A computing device, comprising:
a memory and a processor;
the memory is configured to store computer-executable instructions, and the processor is configured to execute the computer-executable instructions to implement the steps of the business process method of any one of claims 1 to 10.
13. A computer readable storage medium storing computer instructions which, when executed by a processor, implement the steps of the traffic processing method of any of claims 1 to 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110160323.7A CN112988948B (en) | 2021-02-05 | 2021-02-05 | Service processing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110160323.7A CN112988948B (en) | 2021-02-05 | 2021-02-05 | Service processing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112988948A true CN112988948A (en) | 2021-06-18 |
| CN112988948B CN112988948B (en) | 2023-09-19 |
Family
ID=76347646
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110160323.7A Active CN112988948B (en) | 2021-02-05 | 2021-02-05 | Service processing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112988948B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113626573A (en) * | 2021-08-11 | 2021-11-09 | 北京深维智信科技有限公司 | Sales session objection and response extraction method and system |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018223992A1 (en) * | 2017-06-07 | 2018-12-13 | 阿里巴巴集团控股有限公司 | Dialogue generation method and apparatus, and electronic device |
| CN109033270A (en) * | 2018-07-09 | 2018-12-18 | 深圳追科技有限公司 | A method of service knowledge base is constructed based on artificial customer service log automatically |
| CN110019149A (en) * | 2019-01-30 | 2019-07-16 | 阿里巴巴集团控股有限公司 | A kind of method for building up of service knowledge base, device and equipment |
| CN110727764A (en) * | 2019-10-10 | 2020-01-24 | 珠海格力电器股份有限公司 | Phone operation generation method and device and phone operation generation equipment |
| CN110765776A (en) * | 2019-10-11 | 2020-02-07 | 阳光财产保险股份有限公司 | Method and device for generating return visit labeling sample data |
| CN111061847A (en) * | 2019-11-22 | 2020-04-24 | 中国南方电网有限责任公司 | Dialogue generation and corpus expansion method and device, computer equipment and storage medium |
| CN111177350A (en) * | 2019-12-20 | 2020-05-19 | 北京淇瑀信息科技有限公司 | Method, device and system for forming dialect of intelligent voice robot |
| CN111309889A (en) * | 2020-02-27 | 2020-06-19 | 支付宝(杭州)信息技术有限公司 | Method and device for text processing |
| CN111324704A (en) * | 2018-12-14 | 2020-06-23 | 阿里巴巴集团控股有限公司 | Method and device for constructing dialect knowledge base and customer service robot |
| CN111339283A (en) * | 2020-05-15 | 2020-06-26 | 支付宝(杭州)信息技术有限公司 | Method and device for providing customer service answers aiming at user questions |
| CN111611382A (en) * | 2020-05-22 | 2020-09-01 | 贝壳技术有限公司 | Dialect model training method, dialog information generation method, device and system |
| CN111782792A (en) * | 2020-08-05 | 2020-10-16 | 支付宝(杭州)信息技术有限公司 | Method and apparatus for information processing |
| CN111984779A (en) * | 2020-09-10 | 2020-11-24 | 支付宝(杭州)信息技术有限公司 | Dialog text analysis method, device, equipment and readable medium |
-
2021
- 2021-02-05 CN CN202110160323.7A patent/CN112988948B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018223992A1 (en) * | 2017-06-07 | 2018-12-13 | 阿里巴巴集团控股有限公司 | Dialogue generation method and apparatus, and electronic device |
| CN109033270A (en) * | 2018-07-09 | 2018-12-18 | 深圳追科技有限公司 | A method of service knowledge base is constructed based on artificial customer service log automatically |
| CN111324704A (en) * | 2018-12-14 | 2020-06-23 | 阿里巴巴集团控股有限公司 | Method and device for constructing dialect knowledge base and customer service robot |
| CN110019149A (en) * | 2019-01-30 | 2019-07-16 | 阿里巴巴集团控股有限公司 | A kind of method for building up of service knowledge base, device and equipment |
| CN110727764A (en) * | 2019-10-10 | 2020-01-24 | 珠海格力电器股份有限公司 | Phone operation generation method and device and phone operation generation equipment |
| CN110765776A (en) * | 2019-10-11 | 2020-02-07 | 阳光财产保险股份有限公司 | Method and device for generating return visit labeling sample data |
| CN111061847A (en) * | 2019-11-22 | 2020-04-24 | 中国南方电网有限责任公司 | Dialogue generation and corpus expansion method and device, computer equipment and storage medium |
| CN111177350A (en) * | 2019-12-20 | 2020-05-19 | 北京淇瑀信息科技有限公司 | Method, device and system for forming dialect of intelligent voice robot |
| CN111309889A (en) * | 2020-02-27 | 2020-06-19 | 支付宝(杭州)信息技术有限公司 | Method and device for text processing |
| CN111339283A (en) * | 2020-05-15 | 2020-06-26 | 支付宝(杭州)信息技术有限公司 | Method and device for providing customer service answers aiming at user questions |
| CN111611382A (en) * | 2020-05-22 | 2020-09-01 | 贝壳技术有限公司 | Dialect model training method, dialog information generation method, device and system |
| CN111782792A (en) * | 2020-08-05 | 2020-10-16 | 支付宝(杭州)信息技术有限公司 | Method and apparatus for information processing |
| CN111984779A (en) * | 2020-09-10 | 2020-11-24 | 支付宝(杭州)信息技术有限公司 | Dialog text analysis method, device, equipment and readable medium |
Non-Patent Citations (2)
| Title |
|---|
| YUNQI QIU.ET AL: "Stepwise Reasoning for Multi-Relation Question Answering over Knowledge Graph with Weak Supervision", WSDM \'20: PROCEEDINGS OF THE 13TH INTERNATIONAL CONFERENCE ON WEB SEARCH AND DATA MINING, pages 474 * |
| 吕诗宁等: "融合神经网络与电力领域知识的智能客服对话系统研究", 浙江电力, pages 76 - 82 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113626573A (en) * | 2021-08-11 | 2021-11-09 | 北京深维智信科技有限公司 | Sales session objection and response extraction method and system |
| CN113626573B (en) * | 2021-08-11 | 2022-09-27 | 北京深维智信科技有限公司 | Sales session objection and response extraction method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112988948B (en) | 2023-09-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108737667B (en) | Voice quality inspection method and device, computer equipment and storage medium | |
| CN108877839B (en) | Voice quality perception evaluation method and system based on voice semantic recognition technology | |
| KR102353284B1 (en) | Generate audio using neural networks | |
| US11355097B2 (en) | Sample-efficient adaptive text-to-speech | |
| CN110046230B (en) | Method for generating recommended speaking collection, and recommended speaking method and device | |
| CN108597538B (en) | Evaluation method and system of speech synthesis system | |
| US20110144990A1 (en) | Rating speech naturalness of speech utterances based on a plurality of human testers | |
| CN110070452B (en) | Model training method and device, computing equipment and computer readable storage medium | |
| CN110633912A (en) | Method and system for monitoring service quality of service personnel | |
| CN110570208B (en) | Complaint preprocessing method and device | |
| Li et al. | Learning to fool the speaker recognition | |
| Weber et al. | Validations of an alpha version of the E3 Forensic Speech Science System (E3FS3) core software tools | |
| CN112988948A (en) | Service processing method and device | |
| CN108053826A (en) | For the method, apparatus of human-computer interaction, electronic equipment and storage medium | |
| Sharma et al. | Fast Griffin Lim based waveform generation strategy for text-to-speech synthesis | |
| CN110059457B (en) | Body-building method and device | |
| Reddy et al. | Inverse filter based excitation model for HMM‐based speech synthesis system | |
| CN115934904A (en) | Text processing method and device | |
| CN111599342A (en) | Tone selecting method and system | |
| CN114822497A (en) | Method, apparatus, device and medium for training speech synthesis model and speech synthesis | |
| CN113256404A (en) | Data processing method and device | |
| CN113822484A (en) | Service quality evaluation method and device, storage medium and electronic device | |
| CN113421576A (en) | Voice conversion method, device, equipment and storage medium | |
| Shi et al. | SchröWave: Realistic voice generation by solving two-stage conditional Schrödinger bridge problems | |
| Mohíno-Herranz et al. | Precision maximization in anger detection in interactive voice response systems |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20211216 Address after: Room 610, floor 6, No. 618, Wai Road, Huangpu District, Shanghai 200010 Applicant after: Ant Shengxin (Shanghai) Information Technology Co.,Ltd. Address before: 801-11, Section B, 8th floor, 556 Xixi Road, Xihu District, Hangzhou City, Zhejiang Province, 310013 Applicant before: Alipay (Hangzhou) Information Technology Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |