CN112860876A - Session auxiliary processing method and device - Google Patents
Session auxiliary processing method and device Download PDFInfo
- Publication number
- CN112860876A CN112860876A CN202110353074.3A CN202110353074A CN112860876A CN 112860876 A CN112860876 A CN 112860876A CN 202110353074 A CN202110353074 A CN 202110353074A CN 112860876 A CN112860876 A CN 112860876A
- Authority
- CN
- China
- Prior art keywords
- session
- rescue
- sensitive
- conversation
- text
- 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.)
- Pending
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 13
- 238000012545 processing Methods 0.000 claims abstract description 91
- 238000001514 detection method Methods 0.000 claims abstract description 13
- 238000000034 method Methods 0.000 claims description 62
- 238000007781 pre-processing Methods 0.000 claims description 10
- 230000011218 segmentation Effects 0.000 claims description 6
- 238000001356 surgical procedure Methods 0.000 claims 1
- 238000005516 engineering process Methods 0.000 description 15
- 238000004590 computer program Methods 0.000 description 14
- 238000010586 diagram Methods 0.000 description 14
- 230000008569 process Effects 0.000 description 14
- 238000010276 construction Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 7
- 239000013598 vector Substances 0.000 description 6
- 239000003795 chemical substances by application Substances 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 230000018109 developmental process Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 238000004806 packaging method and process Methods 0.000 description 2
- 239000000758 substrate Substances 0.000 description 2
- 206010063659 Aversion Diseases 0.000 description 1
- 238000012369 In process control Methods 0.000 description 1
- 206010044565 Tremor Diseases 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000010965 in-process control Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000036651 mood Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 239000002699 waste material Substances 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
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Computational Linguistics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Health & Medical Sciences (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Telephonic Communication Services (AREA)
Abstract
The present disclosure provides a session auxiliary processing method, including: acquiring audio stream data of a real-time conversation with a user; converting the audio stream data into session texts in real time; detecting the conversation text based on a sensitive word detection algorithm to identify sensitive risk words appearing in the conversation text; under the condition that the sensitive risk words meet preset early warning standards, determining a rescue operation set aiming at the sensitive risk words in a preset rescue operation library; sending a session intervention request and the rescue session set to an auxiliary session end to request the auxiliary session end to access the real-time session, and performing session auxiliary processing based on the rescue session set. The disclosure also provides a conversation auxiliary processing device, an electronic device and a computer storage medium.
Description
Technical Field
The present disclosure relates to the field of big data technologies, and in particular, to a method and an apparatus for session assistance processing.
Background
With the rapid development of the financial industry and other industries, the form of user service is in diversified development. As a common user service form, the outbound session can provide efficient consultation services to a large number of users and perform convenient product outbound recommendation to a large number of users.
In the process of implementing the technical scheme, the inventor finds that due to the fact that the mobility of the outbound seat personnel is strong, in the process of real-time communication between the outbound seat personnel and the user, the problems of unsatisfactory conversation effect and high user complaint rate often occur due to lack of professional knowledge or insufficient processing experience of the outbound seat personnel.
Disclosure of Invention
One aspect of the present disclosure provides a session assistance processing method, including: acquiring audio stream data of a real-time conversation with a user; converting the audio stream data into session texts in real time; detecting the conversation text based on a sensitive word detection algorithm to identify sensitive risk words appearing in the conversation text; under the condition that the sensitive risk words meet preset early warning standards, determining a rescue operation set aiming at the sensitive risk words in a preset rescue operation library; sending a session intervention request and the rescue session set to an auxiliary session end to request the auxiliary session end to access the real-time session, and performing session auxiliary processing based on the rescue session set.
Optionally, the detecting the conversation text based on the sensitive word detection algorithm to identify the sensitive risk word appearing in the conversation text includes: preprocessing the conversation text to obtain a keyword set in the conversation text, wherein the preprocessing at least comprises word segmentation, stop word removal and punctuation mark removal; and determining the keywords with the similarity higher than a preset threshold value with the sensitive characteristic words by using a similarity matching algorithm based on a preset sensitive characteristic word bank to serve as the sensitive risk words.
Optionally, the sensitive feature word stock construction method includes: acquiring audio stream data of a complaint session; converting audio stream data of the complaint session into complaint session text; and extracting key words with word frequency characteristics meeting preset conditions from the complained conversation text to form the sensitive characteristic word bank.
Optionally, the method for determining whether the sensitive risk word meets a preset early warning standard includes: calculating a risk score of the conversation text according to the sensitive characteristic words associated with the sensitive risk words and the preset weight associated with the sensitive characteristic words; and under the condition that the risk score reaches a preset threshold value, determining that the sensitive risk word meets the early warning standard.
Optionally, the construction method of the rescue operation library includes: acquiring audio stream data of the rescue session; converting audio stream data of the rescue session into a rescue session text; extracting at least one rescue operation with the occurrence frequency higher than a preset threshold value from the rescue session text; determining a category label associated with each of the rescue procedures according to preset classification criteria; and summarizing the at least one rescue operation based on the category label to obtain the rescue operation library.
Optionally, the determining a set of salvages for the sensitive risk word in a preset salvages library includes: determining a target category label matched with the risk sensitive word; determining a preset number of rescue operations with highest recommendation weight in the rescue operation set associated with the target category label to form the rescue operation set aiming at the sensitive risk word, wherein the recommendation weight associated with any rescue operation is determined by the occurrence frequency of the rescue operation in the rescue conversation text.
Optionally, the method further comprises: sending the latest conversation sub-text of a preset turn in the conversation text to the auxiliary conversation end; the sending the rescue operation set to a preset auxiliary session end comprises: and sending a rescue operation set sorted according to the recommended weight to the auxiliary session end so that the auxiliary session end carries out session auxiliary processing based on the latest session sub-text of the preset turn and the rescue operation set sorted according to the recommended weight.
Another aspect of the present disclosure provides a conversation assistance processing apparatus, including: the acquisition module is used for acquiring audio stream data of real-time conversation with a user; the first processing module is used for converting the audio stream data into a conversation text in real time; the second processing module is used for detecting the conversation text based on a sensitive word detection algorithm so as to identify sensitive risk words appearing in the conversation text; the third processing module is used for determining a rescue operation set aiming at the sensitive risk words in a preset rescue operation library under the condition that the sensitive risk words meet preset early warning standards; and the fourth processing module is used for sending a session intervention request and the rescue operation set to an auxiliary session end so as to request the auxiliary session end to access the real-time session, and performing session auxiliary processing based on the rescue operation set.
Optionally, the second processing module includes: the first processing submodule is used for preprocessing the conversation text to obtain a keyword set in the conversation text, wherein the preprocessing at least comprises word segmentation, stop word removal and punctuation mark removal; and the second processing sub-module is used for determining the keywords with the similarity higher than a preset threshold value with the sensitive characteristic words by utilizing a similarity matching algorithm based on a preset sensitive characteristic word bank to serve as the sensitive risk words.
Optionally, the sensitive feature word stock construction method includes: acquiring audio stream data of a complaint session; converting audio stream data of the complaint session into complaint session text; and extracting key words with word frequency characteristics meeting preset conditions from the complained conversation text to form the sensitive characteristic word bank.
Optionally, the third processing module is further configured to: calculating a risk score of the conversation text according to the sensitive characteristic words associated with the sensitive risk words and the preset weight associated with the sensitive characteristic words; and under the condition that the risk score reaches a preset threshold value, determining that the sensitive risk word meets the early warning standard.
Optionally, the construction method of the rescue operation library includes: acquiring audio stream data of the rescue session; converting audio stream data of the rescue session into a rescue session text; extracting at least one rescue operation with the occurrence frequency higher than a preset threshold value from the rescue session text; determining a category label associated with each of the rescue procedures according to preset classification criteria; and summarizing the at least one rescue operation based on the category label to obtain the rescue operation library.
Optionally, the third processing module includes: the third processing submodule is used for determining a target category label matched with the risk sensitive word; and the fourth processing submodule is used for determining a preset number of rescue operations with the highest recommendation weight in the rescue operation set associated with the target class label to form a rescue operation set aiming at the sensitive risk words, wherein the recommendation weight associated with any rescue operation is determined by the occurrence frequency of the rescue operation in the rescue session text.
Optionally, the fourth processing module is further configured to: sending the latest conversation sub-text of a preset turn in the conversation text to the auxiliary conversation end; and sending a rescue conversation set sorted according to the recommended weight to the auxiliary conversation end through a fourth processing module so that the auxiliary conversation end carries out conversation auxiliary processing based on the latest conversation sub-text of the preset turn and the rescue conversation set sorted according to the recommended weight.
Another aspect of the present disclosure provides an electronic device comprising one or more processors; memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the methods of embodiments of the present disclosure.
Another aspect of the present disclosure provides a computer-readable storage medium storing computer-executable instructions that, when executed, implement the method of embodiments of the present disclosure.
Another aspect of the present disclosure provides a computer program product comprising computer readable instructions, wherein the computer readable instructions are configured to perform a session assistance processing method of an embodiment of the present disclosure when executed.
Drawings
For a more complete understanding of the present disclosure and the advantages thereof, reference is now made to the following descriptions taken in conjunction with the accompanying drawings, in which,
fig. 1 schematically illustrates a system architecture of a session assistance processing method and apparatus according to an embodiment of the present disclosure;
FIG. 2 schematically illustrates a flow chart of a method of session assistance processing according to an embodiment of the present disclosure;
FIG. 3 schematically illustrates a flow chart of another session assistance processing method according to an embodiment of the present disclosure;
FIG. 4 is a schematic diagram that schematically illustrates an auxiliary session-side display interface, in accordance with an embodiment of the present disclosure;
fig. 5 schematically illustrates a block diagram of a conversation assistance processing apparatus according to an embodiment of the present disclosure;
fig. 6 schematically shows a block diagram of an electronic device according to an embodiment of the disclosure.
Detailed Description
Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It is to be understood that such description is merely illustrative and not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, operations, and/or components, but do not preclude the presence or addition of one or more other features, operations, or components.
All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.
Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B and C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).
Some block diagrams and/or flow diagrams are shown in the figures. It will be understood that some blocks of the block diagrams and/or flowchart illustrations, or combinations thereof, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable session assistance processing apparatus such that the instructions, which execute via the processor, create means for implementing the functions/acts specified in the block diagrams and/or flowchart block or blocks. The techniques of this disclosure may be implemented in hardware and/or software (including firmware, microcode, etc.). In addition, the techniques of this disclosure may take the form of a computer program product on a computer-readable storage medium having instructions stored thereon for use by or in connection with an instruction execution system.
The embodiment of the disclosure provides a conversation assistance processing method and a processing device capable of applying the method. The method specifically comprises the following operations of firstly obtaining audio stream data of a real-time conversation with a user, converting the audio stream data into a conversation text in real time, then detecting the conversation text based on a sensitive word detection algorithm to identify sensitive risk words appearing in the conversation text, next, determining a rescue conversation set aiming at the sensitive risk words in a preset rescue conversation library under the condition that the sensitive risk words meet preset early warning standards, finally, sending a conversation intervention request and the rescue conversation set to an auxiliary conversation end to request the auxiliary conversation end to access the real-time conversation, and carrying out conversation auxiliary processing based on the rescue conversation set.
Fig. 1 schematically shows a system architecture of a session assistance processing method and apparatus according to an embodiment of the present disclosure. It should be noted that fig. 1 is only an example of a system architecture to which the embodiments of the present disclosure may be applied to help those skilled in the art understand the technical content of the present disclosure, and does not mean that the embodiments of the present disclosure may not be applied to other devices, systems, environments or scenarios.
As shown in fig. 1, the system architecture 100 includes a real-time session end 102, a processor 103, and an auxiliary session end 104, where the real-time session end 102 is configured to perform a real-time session with a user (e.g., a user 101 in the figure), and the processor 103 is configured to obtain audio stream data of the real-time session between the real-time session end 102 and the user, and determine whether the real-time session reaches a preset early warning standard based on the audio stream data, and if so, request the auxiliary session end 104 to access the real-time session and perform session auxiliary processing.
Specifically, the processor 103 obtains audio stream data of a real-time conversation between the real-time conversation terminal 102 and a user, converts the audio stream data into a conversation text in real time, detects the conversation text based on a sensitive word detection algorithm to identify a sensitive risk word appearing in the conversation text, determines a rescue conversation set for the sensitive risk word in a preset rescue conversation library under the condition that the sensitive risk word meets a preset early warning standard, and finally sends a conversation intervention request and a rescue conversation set to the auxiliary conversation terminal 104 to request the auxiliary conversation terminal 104 to access the real-time conversation and perform conversation auxiliary processing based on the rescue conversation set.
It should be noted that the session assistance processing method and apparatus in the embodiments of the present disclosure may be used in the financial field, and may also be used in any field other than the financial field. The present disclosure will be described in detail below with reference to the drawings and specific embodiments.
Fig. 2 schematically shows a flowchart of a session assistance processing method according to an embodiment of the present disclosure, and as shown in fig. 2, the method 200 may include operations S210 to S250.
In operation S210, audio stream data for a real-time conversation with a user is acquired.
In this embodiment, specifically, audio stream data of a real-time session with a user is obtained, where the real-time session may specifically be real-time call data in a customer service reception process, and the customer service reception process may include, for example, a business consultation response process, an after-sales service processing process, an opinion complaint reception process, a product outbound recommendation process, and the like. And acquiring real-time call data between the real-time conversation end and the user, and detecting sensitive risk words on the real-time call data, wherein the sensitive risk words indicate that the real-time call quality inspection is unqualified and the possibility of complaint by the user exists.
Under the condition that the real-time conversation is judged to reach the preset early warning standard based on the sensitive risk words, the auxiliary conversation end is requested to access the real-time conversation for in-process control, so that the complaint sensitive events can be solved in time, the complaint rate of the clients after the event is effectively reduced, and the customer service reception experience of the users is effectively improved. In practical application, the real-time conversation end is generally a foreground operator conversation end, and foreground operators have strong mobility and may be deficient in professional skill level or emergency treatment experience. The auxiliary conversation end can be a background agent personnel conversation end with high professional skill level and rich emergency treatment experience, or can also be a high-job staff conversation end, such as a client manager conversation end.
Next, the audio stream data is converted into a conversation text in real time in operation S220.
In this embodiment, specifically, an Automatic Speech Recognition (ASR) technology is used to perform real-time text conversion processing on the acquired audio stream data, so as to obtain a session list between the foreground personnel and the user. Further, separating the agent sentences and the user sentences in the conversation list to obtain agent conversation texts and user conversation texts.
Next, in operation S230, the conversation text is detected based on a sensitive word detection algorithm to identify sensitive risk words appearing in the conversation text.
In this embodiment, specifically, the session text is preprocessed to obtain a keyword set in the session text, where the preprocessing may include, for example, word segmentation, stop word removal, punctuation removal, and the like. Stop words are words which are automatically filtered out in the process of processing natural language data in order to save storage space and improve retrieval efficiency in the process of information retrieval, and the stop words can comprise, for example, mood auxiliary words, adverbs, prepositions, conjunctions and the like, and exemplarily stop words such as 'in' and 'on' and the like which appear in a conversation text are removed.
And determining the keywords with the similarity higher than a preset threshold value with the sensitive characteristic words by using a similarity matching algorithm based on a preset sensitive characteristic word bank to serve as sensitive risk words. The sensitive risk words indicate that the real-time conversation has the possibility of being complained afterwards, and intervention auxiliary processing needs to be carried out on the real-time conversation so as to reduce the complaint rate of the post-event user through in-event control.
The construction method of the sensitive feature word stock comprises the following steps: acquiring audio stream data of a complaint session; converting audio stream data of the complaint session into complaint session text; and extracting key words with word frequency characteristics meeting preset conditions from the complained session text to form a sensitive characteristic word bank.
Specifically, complaint records of past users for the outbound service are collected, and audio stream data of a complaint session is acquired based on the content of a complaint number, customer information, an agent number, and the like described in the complaint records. And performing text conversion processing on the audio stream data to obtain a complained session text, and performing preprocessing such as word segmentation, stop word removal, punctuation mark removal and the like on the complained session text to obtain a keyword set in the complained session text.
And then, extracting keywords with word frequency characteristics meeting preset conditions from a keyword set in the complained session text to form a sensitive characteristic word bank. The word frequency characteristics indicate the importance degree and the distinguishing capability of the keywords in the text of the complained conversation, and illustratively, the word frequency characteristics of each keyword are represented by using TF-IDF values. Specifically, TF-IDF is TF × IDF, TF is T/T, and IDF is N/(N +1), where TF denotes a word frequency, IDF denotes a reverse document frequency, T denotes a frequency of occurrence of a certain keyword in a certain complaint session text, T denotes a total number of words of the complaint session text, N denotes a total number of acquired complaint session texts, and N denotes a total number of complaint session texts including the keyword. The larger the TF-IDF value is, the higher the importance degree of the key word is represented, the stronger the distinguishing capability is, and the higher the text of the complained conversation can be represented. And extracting key words with TF-IDF values larger than a preset threshold value in the complained session text to form a sensitive feature word bank.
Next, in operation S240, in case that the sensitive risk word satisfies the preset early warning criterion, a rescue set for the sensitive risk word is determined in a preset rescue library.
In this embodiment, specifically, since the risk levels of different sensitive feature words are different, different weight values may be preset for different sensitive feature words. And calculating the risk score of the conversation text according to the sensitive characteristic words associated with the sensitive risk words and the preset weight associated with the sensitive characteristic words. And under the condition that the risk score reaches a preset threshold value, determining that the sensitive risk word meets a preset early warning standard.
And under the condition that the sensitive risk words meet the preset early warning standard, determining a rescue operation set aiming at the sensitive risk words in a preset rescue operation library. Specifically, the rescue words matched with the sensitive risk words are determined in the rescue word library so as to reduce the probability of the user's post-event complaints by delivering the rescue words to the user.
The construction method of the rescue operation library comprises the following steps: acquiring audio stream data of the rescue session; converting audio stream data of the rescue session into a rescue session text; extracting at least one rescue operation with the occurrence frequency higher than a preset threshold value from the rescue session text; determining category labels associated with the rescue operations according to preset classification criteria; and summarizing at least one rescue operation based on the category label to obtain a rescue operation library.
Specifically, based on complaint records of past users for outbound service, rescue session records with good subsequent resolution effect are collected. And acquiring audio stream data of the rescue session based on contents such as a return visit number, client information, an agent number and the like recorded in the rescue session record. The audio stream data of the rescue conversation is subjected to text transferring processing to obtain a rescue conversation text, and the acquired audio stream data of the rescue conversation may exist in multiple numbers, so that a rescue conversation text set containing multiple rescue conversation texts may be obtained after the text transferring processing. And determining at least one rescue operation with the occurrence frequency higher than a preset threshold value in the text set of the rescue session to form a rescue operation library. The higher the occurrence frequency of the rescue operation, the more remarkable the rescue effect of the characterized rescue operation is, and the higher the importance degree is, the more worth being recommended the use of the rescue operation.
The source scenes of the sensitive feature words can be generally classified into types of service consultation scenes, service handling scenes, call-out recommendation scenes and the like, the sensitive feature words in the service consultation scenes can include words with bad attitudes, slow response, untimely response, tremble, complaint, unfamiliarity, time waste and the like, the sensitive feature words in the service handling scenes can include words with privacy disclosure, violation, delay, overdue, errors, verification failure, repeated submission and the like, and the sensitive feature words in the call-out recommendation scenes can include words with disturbance, unneeded, aversion, influence, repetition and the like. The rescue techniques adapted to the sensitive feature words conform to the source scenes of the sensitive feature words, so that the applicable source scenes can be used as preset classification standards, and the class labels associated with the rescue techniques are determined based on the preset classification standards. And summarizing at least one rescue operation based on the category labels to obtain a rescue operation library.
Next, in operation S250, a session intervention request and a rescue session set are sent to the auxiliary session end to request the auxiliary session end to access the real-time session, and a session assistance process is performed based on the rescue session set.
In this embodiment, specifically, a session intervention request is sent to the auxiliary session end to request the auxiliary session end to access the real-time session. Meanwhile, a rescue conversation set aiming at the sensitive risk words is sent to the auxiliary conversation end, so that the auxiliary conversation end carries out conversation auxiliary processing on the basis of the rescue conversation set. By delivering a rescue conversation to the user, the conversation which is possibly developed into a complaint event is subjected to in-flight processing, which can effectively reduce the incidence rate of actual complaints in the outbound service. Optionally, the latest conversation sub-text of the preset turn in the conversation text may also be sent to the auxiliary conversation end, so that the auxiliary conversation end grasps the user interaction intention and selects or provides an applicable rescue operation by itself based on the latest conversation sub-text of the preset turn.
According to the embodiment of the disclosure, audio stream data of a real-time conversation with a user is acquired; converting audio stream data into a conversation text in real time; detecting the conversation text based on a sensitive word detection algorithm to identify sensitive risk words appearing in the conversation text; determining a rescue operation set aiming at the sensitive risk words in a preset rescue operation library under the condition that the sensitive risk words meet preset early warning standards; and sending a session intervention request and a rescue conversation set to the auxiliary conversation end to request the auxiliary conversation end to access the real-time session, and performing session auxiliary processing based on the rescue conversation set. The method comprises the steps of identifying sensitive risk words appearing in a session text, determining a rescue technology set aiming at the sensitive risk words in a preset rescue technology library under the condition that the sensitive risk words meet preset early warning standards, requesting an auxiliary session end to access a real-time session, and performing session auxiliary processing based on the rescue technology set. By introducing a natural language processing technology and a big data analysis technology, the real-time conversation between the outbound seat and the user is effectively monitored, the real-time conversation which is possibly developed into a complaint event is processed in the process, the incidence rate of actual complaints in the outbound service can be effectively reduced, the service quality of the financial outbound service is favorably improved, and the communication efficiency between the outbound seat and the user is improved. In addition, the method is beneficial to controlling the training cost of the outbound seat personnel, shortens the training time and can well meet the rapid development requirement of financial business.
Optionally, after the sensitive feature word bank is constructed, the audio stream data of the complaint session that has evolved into the complaint event can be fully utilized, and the sensitive feature word bank is updated by performing fixed-point analysis on the audio stream data of the complaint session, specifically including updating the sensitive feature words in the sensitive feature word bank and updating the preset weights associated with the sensitive feature words. Similarly, after the rescue operation library is constructed, the audio stream data of the rescue conversation with good solution effect can be fully utilized, and the rescue operation library is updated by carrying out fixed-point analysis on the audio stream data of the rescue conversation.
Fig. 3 schematically illustrates a flowchart of another session assistance processing method according to an embodiment of the present disclosure, and as shown in fig. 3, the method 300 may include operations S210 to S220, S310 to S330, and S250.
In operation S210, audio stream data for a real-time conversation with a user is acquired.
Next, the audio stream data is converted into a conversation text in real time in operation S220.
Next, in operation S310, the session text is preprocessed to obtain a keyword set in the session text, and keywords with similarity higher than a preset threshold with the sensitive feature word are determined by using a similarity matching algorithm based on a preset sensitive feature word library to serve as sensitive risk words.
In this embodiment, specifically, after the session text is preprocessed, a keyword set in the session text is obtained. And determining key words with the similarity higher than a preset threshold value with the sensitive characteristic words based on a preset sensitive characteristic word bank by using a similarity matching algorithm to serve as sensitive risk words in the conversation text. Or, searching whether a sensitive feature word matched with each keyword exists in the sensitive feature word library, and if yes, taking the corresponding keyword as a sensitive risk word in the conversation text. Optionally, each keyword in the session text may be converted into a first feature vector, and each sensitive feature word in the sensitive feature word bank may be converted into a second feature vector by a word embedding method. And performing similarity matching on each first feature vector and each second feature vector according to a cosine similarity algorithm, judging that the keywords corresponding to two feature vectors are similar to the sensitive feature words in semantics when the similarity of the two feature vectors is higher than a preset threshold value, and taking the keywords as the sensitive risk words in the session text.
Next, in operation S320, a risk score of the conversation text is calculated according to the sensitive feature words associated with the sensitive risk words and according to the preset weights associated with the sensitive feature words, and it is determined that the sensitive risk words satisfy a preset early warning standard when the risk score reaches a preset threshold.
In this embodiment, specifically, the risk levels of different sensitive feature words are different, and therefore, the weight values pre-assigned to the sensitive feature words may be different, and the weight values pre-assigned to the different sensitive feature words may be set reasonably according to the complaint history in the field of the outbound service. For example, the sensitive feature word "violation" may be at a higher risk level than the sensitive feature word "disturbance" and, therefore, a higher weight value may be preset for the sensitive feature word "violation". Different sensitive characteristic words are gathered in the sensitive characteristic word bank, and preset weights associated with the sensitive characteristic words are also gathered. By setting the sensitive feature word bank, sensitive risk words appearing in the conversation text can be conveniently identified, and risk scores of the conversation text can be conveniently calculated based on the sensitive risk words.
Next, in operation S330, in a case that the sensitive risk word meets a preset early warning criterion, a target category label matching the risk sensitive word is determined, and a preset number of rescue techniques with the highest recommendation weight are determined in a rescue technique set associated with the target category label to form a rescue technique set for the sensitive risk word.
In this embodiment, specifically, the sensitive feature words generally have associated source scenario features, and the source scenarios include, for example, types such as a business consultation scenario, a business transaction scenario, and an outbound recommendation scenario. Therefore, the target category label matched with the sensitive risk word can be determined according to the source scene characteristics of the sensitive risk word. Determining a set of rescue operations associated with the target category label in the rescue operations library, and determining a preset number of rescue operations with highest recommendation weight in the set of rescue operations, constituting a set of rescue operations for the sensitive risk word.
Illustratively, for the sensitive risk words "authentication failure" and "repeat submission" in the business handling class scene, the target class label matched with the sensitive risk words is determined as "appeasing in the business handling class scene + solving the call", and the rescue call matched with the sensitive risk words is determined to include "you are in urgency, here, help you check the situation immediately, and trouble you provide a single number, good.
The recommendation weight associated with any rescue operation is determined by the frequency of occurrence of the rescue operation in the text of the rescue session, the higher the frequency of occurrence, the more times the user complaints are effectively resolved by the characterization of the use of the rescue operation, the more positive the effect of the operation is, and therefore, the higher the recommendation weight for the rescue operation is.
Next, in operation S250, a session intervention request and a rescue session set are sent to the auxiliary session end to request the auxiliary session end to access the real-time session, and a session assistance process is performed based on the rescue session set.
In this embodiment, specifically, a session intervention request is sent to the auxiliary session end to request the auxiliary session end to access the real-time session for session assistance processing. And simultaneously, sending a rescue operation set ordered according to the recommended weight to the auxiliary session end so that the auxiliary session end performs real-time session auxiliary processing by taking the rescue operation set as a reference and combining with the latest session sub-text of a preset turn in the session text. Through the conversation auxiliary mode, the problem that the real-time conversation is possibly carried out into a complaint event is solved in the affairs, the incidence rate of actual complaints in the outbound service can be effectively reduced, the communication efficiency between the outbound seat and the user is favorably improved, and meanwhile, the interactive experience of the user in the financial outbound service is favorably improved.
Fig. 4 schematically shows a schematic diagram of a display interface of an auxiliary session end according to an embodiment of the present disclosure, where a session intervention request, a latest session sub-text of a preset turn in a session text, a service keyword and a risk sensitive word in the session text, and a rescue operation set sorted according to a recommended weight are displayed in the display interface, so that the auxiliary session end accesses a real-time session between an outbound seat and a user and performs session assistance processing based on the latest session sub-text and the rescue operation set of the preset turn.
Fig. 5 schematically shows a block diagram of a conversation assistance processing apparatus according to an embodiment of the present disclosure.
As shown in fig. 5, the apparatus 500 includes an obtaining module 501, a first processing module 502, a second processing module 503, a third processing module 504, and a fourth processing module 505.
Specifically, the obtaining module 501 is configured to obtain audio stream data of a real-time session with a user; a first processing module 502, configured to convert audio stream data into a session text in real time; the second processing module 503 is configured to detect the session text based on a sensitive word detection algorithm to identify a sensitive risk word appearing in the session text; a third processing module 504, configured to determine, in the preset rescue technology library, a rescue technology set for the sensitive risk word when the sensitive risk word meets a preset early warning criterion; a fourth processing module 505, configured to send a session intervention request and a rescue session set to the auxiliary session end to request the auxiliary session end to access the real-time session, and perform session auxiliary processing based on the rescue session set.
According to the embodiment of the disclosure, audio stream data of a real-time conversation with a user is acquired; converting audio stream data into a conversation text in real time; detecting the conversation text based on a sensitive word detection algorithm to identify sensitive risk words appearing in the conversation text; determining a rescue operation set aiming at the sensitive risk words in a preset rescue operation library under the condition that the sensitive risk words meet preset early warning standards; and sending a session intervention request and a rescue conversation set to the auxiliary conversation end to request the auxiliary conversation end to access the real-time session, and performing session auxiliary processing based on the rescue conversation set. The method comprises the steps of identifying sensitive risk words appearing in a session text, determining a rescue technology set aiming at the sensitive risk words in a preset rescue technology library under the condition that the sensitive risk words meet preset early warning standards, requesting an auxiliary session end to access a real-time session, and performing session auxiliary processing based on the rescue technology set. By introducing a natural language processing technology and a big data analysis technology, the real-time conversation between the outbound seat and the user is effectively monitored, the real-time conversation which is possibly developed into a complaint event is processed in the process, the incidence rate of actual complaints in the outbound service can be effectively reduced, the service quality of the financial outbound service is favorably improved, and the communication efficiency between the outbound seat and the user is improved. In addition, the method is beneficial to controlling the training cost of the outbound seat personnel, shortens the training time and can well meet the rapid development requirement of financial business.
As a possible embodiment, the second processing module includes: the first processing submodule is used for preprocessing the conversation text to obtain a keyword set in the conversation text, wherein the preprocessing at least comprises word segmentation, stop word removal and punctuation mark removal; and the second processing sub-module is used for determining the keywords with the similarity higher than a preset threshold value with the sensitive characteristic words by utilizing a similarity matching algorithm based on a preset sensitive characteristic word bank to serve as the sensitive risk words.
As a possible embodiment, the method for constructing the sensitive feature word bank includes: acquiring audio stream data of a complaint session; converting audio stream data of the complaint session into complaint session text; and extracting key words with word frequency characteristics meeting preset conditions from the complained session text to form a sensitive characteristic word bank.
As a possible embodiment, the third processing module is further configured to: calculating a risk score of the conversation text according to the sensitive characteristic words associated with the sensitive risk words and the preset weights associated with the sensitive characteristic words; and determining that the sensitive risk word meets the early warning standard under the condition that the risk score reaches a preset threshold value.
As a possible embodiment, the construction method of the salvage technology library includes: acquiring audio stream data of the rescue session; converting audio stream data of the rescue session into a rescue session text; extracting at least one rescue operation with the occurrence frequency higher than a preset threshold value from the rescue session text; determining category labels associated with the rescue operations according to preset classification criteria; and summarizing at least one rescue operation based on the category label to obtain a rescue operation library.
As a possible embodiment, the third processing module includes: the third processing submodule is used for determining a target category label matched with the risk sensitive word; and the fourth processing submodule is used for determining a preset number of rescue operations with the highest recommendation weight in the rescue operation set associated with the target class label to form a rescue operation set aiming at the sensitive risk words, wherein the recommendation weight associated with any rescue operation is determined by the occurrence frequency of the rescue operation in the rescue session text.
As a possible embodiment, the fourth processing module is further configured to: the latest conversation sub-text of the preset turn in the conversation text is sent to the auxiliary conversation end; and sending the rescue operation set sorted according to the recommended weight to the auxiliary conversation end through a fourth processing module so that the auxiliary conversation end carries out conversation auxiliary processing on the basis of the latest conversation sub-text of the preset turn and the rescue operation set sorted according to the recommended weight.
It should be noted that, in the embodiments of the present disclosure, the implementation of the apparatus portion is the same as or similar to the implementation of the method portion, and is not described herein again.
Any of the modules according to embodiments of the present disclosure, or at least part of the functionality of any of them, may be implemented in one module. Any one or more of the modules according to the embodiments of the present disclosure may be implemented by being split into a plurality of modules. Any one or more of the modules according to the embodiments of the present disclosure may be implemented at least in part as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in any other reasonable manner of hardware or firmware by integrating or packaging the circuit, or in any one of three implementations, or in any suitable combination of any of the software, hardware, and firmware. Or one or more of the modules according to embodiments of the disclosure, may be implemented at least partly as computer program modules which, when executed, may perform corresponding functions.
For example, any number of the obtaining module 501, the first processing module 502, the second processing module 503, the third processing module 504 and the fourth processing module 505 may be combined and implemented in one module, or any one of the modules may be split into a plurality of modules. Alternatively, at least part of the functionality of one or more of these modules may be combined with at least part of the functionality of the other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the obtaining module 501, the first processing module 502, the second processing module 503, the third processing module 504, and the fourth processing module 505 may be at least partially implemented as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented by hardware or firmware in any other reasonable manner of integrating or packaging a circuit, or implemented by any one of three implementations of software, hardware, and firmware, or by a suitable combination of any several of them. At least one of the obtaining module 501, the first processing module 502, the second processing module 503, the third processing module 504 and the fourth processing module 505 may be at least partly implemented as a computer program module, which when executed may perform a corresponding function.
Fig. 6 schematically shows a block diagram of an electronic device according to an embodiment of the disclosure. The electronic device shown in fig. 6 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.
As shown in fig. 6, the electronic device 600 includes a processor 610, a computer-readable storage medium 620. The electronic device 600 may perform a method according to an embodiment of the present disclosure.
In particular, the processor 610 may comprise, for example, a general purpose microprocessor, an instruction set processor and/or related chip set and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), or the like. The processor 610 may also include onboard memory for caching purposes. The processor 610 may be a single processing module or a plurality of processing modules for performing different actions of a method flow according to embodiments of the disclosure.
Computer-readable storage medium 620, for example, may be a non-volatile computer-readable storage medium, specific examples including, but not limited to: magnetic storage devices, such as magnetic tape or Hard Disk Drives (HDDs); optical storage devices, such as compact disks (CD-ROMs); a memory, such as a Random Access Memory (RAM) or a flash memory; and so on.
The computer-readable storage medium 620 may include a computer program 621, which computer program 621 may include code/computer-executable instructions that, when executed by the processor 610, cause the processor 610 to perform a method according to an embodiment of the disclosure, or any variation thereof.
The computer program 621 may be configured with, for example, computer program code comprising computer program modules. For example, in an example embodiment, code in computer program 621 may include one or more program modules, including, for example, module 621A, module 621B. It should be noted that the division and number of the modules are not fixed, and those skilled in the art may use suitable program modules or program module combinations according to actual situations, so that the processor 610 may execute the method according to the embodiment of the present disclosure or any variation thereof when the program modules are executed by the processor 610.
According to an embodiment of the present disclosure, at least one of the obtaining module 501, the first processing module 502, the second processing module 503, the third processing module 504 and the fourth processing module 505 may be implemented as a computer program module described with reference to fig. 6, which, when executed by the processor 610, may implement the respective operations described above.
The present disclosure also provides a computer-readable storage medium, which may be contained in the apparatus/device/system described in the above embodiments; or may exist separately and not be assembled into the device/apparatus/system. The computer-readable storage medium carries one or more programs which, when executed, implement the method according to an embodiment of the disclosure.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
It will be understood by those skilled in the art that while the present disclosure has been shown and described with reference to certain exemplary embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present disclosure as defined by the appended claims and their equivalents. Accordingly, the scope of the present disclosure should not be limited to the above-described embodiments, but should be defined not only by the appended claims, but also by equivalents thereof.
Claims (10)
1. A conversation assistance processing method comprises the following steps:
acquiring audio stream data of a real-time conversation with a user;
converting the audio stream data into session texts in real time;
detecting the conversation text based on a sensitive word detection algorithm to identify sensitive risk words appearing in the conversation text;
under the condition that the sensitive risk words meet preset early warning standards, determining a rescue operation set aiming at the sensitive risk words in a preset rescue operation library;
sending a session intervention request and the rescue session set to an auxiliary session end to request the auxiliary session end to access the real-time session, and performing session auxiliary processing based on the rescue session set.
2. The method of claim 1, wherein the detecting the conversation text based on the sensitive word detection algorithm to identify the sensitive risk words present in the conversation text comprises:
preprocessing the conversation text to obtain a keyword set in the conversation text, wherein the preprocessing at least comprises word segmentation, stop word removal and punctuation mark removal;
and determining the keywords with the similarity higher than a preset threshold value with the sensitive characteristic words by using a similarity matching algorithm based on a preset sensitive characteristic word bank to serve as the sensitive risk words.
3. The method of claim 2, wherein the sensitive feature word bank is constructed by a method comprising:
acquiring audio stream data of a complaint session;
converting audio stream data of the complaint session into complaint session text;
and extracting key words with word frequency characteristics meeting preset conditions from the complained conversation text to form the sensitive characteristic word bank.
4. The method of claim 2, wherein the method of determining whether the sensitive risk word satisfies a preset pre-warning criterion comprises:
calculating a risk score of the conversation text according to the sensitive characteristic words associated with the sensitive risk words and the preset weight associated with the sensitive characteristic words;
and under the condition that the risk score reaches a preset threshold value, determining that the sensitive risk word meets the early warning standard.
5. The method of claim 1, wherein the salvage surgery library is constructed by a method comprising:
acquiring audio stream data of the rescue session;
converting audio stream data of the rescue session into a rescue session text;
extracting at least one rescue operation with the occurrence frequency higher than a preset threshold value from the rescue session text;
determining a category label associated with each of the rescue procedures according to preset classification criteria;
and summarizing the at least one rescue operation based on the category label to obtain the rescue operation library.
6. The method of claim 5, wherein the determining a rescue set for the sensitive risk word in a preset rescue library comprises:
determining a target category label matched with the risk sensitive word;
determining a preset number of rescue techniques with highest recommendation weight in the set of rescue techniques associated with the target category label to constitute a set of rescue techniques for the sensitive risk word,
wherein the recommendation weight associated with any of the rescue sessions is determined by the frequency of occurrence of the rescue session in the rescue session text.
7. The method of claim 1, further comprising:
sending the latest conversation sub-text of a preset turn in the conversation text to the auxiliary conversation end;
the sending the rescue operation set to a preset auxiliary session end comprises:
and sending a rescue operation set sorted according to the recommended weight to the auxiliary session end so that the auxiliary session end carries out session auxiliary processing based on the latest session sub-text of the preset turn and the rescue operation set sorted according to the recommended weight.
8. A conversation assistance processing apparatus comprising:
the acquisition module is used for acquiring audio stream data of real-time conversation with a user;
the first processing module is used for converting the audio stream data into a conversation text in real time;
the second processing module is used for detecting the conversation text based on a sensitive word detection algorithm so as to identify sensitive risk words appearing in the conversation text;
the third processing module is used for determining a rescue operation set aiming at the sensitive risk words in a preset rescue operation library under the condition that the sensitive risk words meet preset early warning standards;
and the fourth processing module is used for sending a session intervention request and the rescue operation set to an auxiliary session end so as to request the auxiliary session end to access the real-time session, and performing session auxiliary processing based on the rescue operation set.
9. An electronic device, comprising:
one or more processors; and
a memory for storing one or more programs,
wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-7.
10. A computer readable storage medium having stored thereon executable instructions which, when executed by a processor, cause the processor to carry out the method of any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110353074.3A CN112860876A (en) | 2021-03-31 | 2021-03-31 | Session auxiliary processing method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110353074.3A CN112860876A (en) | 2021-03-31 | 2021-03-31 | Session auxiliary processing method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112860876A true CN112860876A (en) | 2021-05-28 |
Family
ID=75992024
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110353074.3A Pending CN112860876A (en) | 2021-03-31 | 2021-03-31 | Session auxiliary processing method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112860876A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113449350A (en) * | 2021-06-30 | 2021-09-28 | 中国工商银行股份有限公司 | Management method, device, equipment and medium for USB outgoing sensitive information |
CN113822062A (en) * | 2021-09-22 | 2021-12-21 | 未鲲(上海)科技服务有限公司 | Text data processing method, device and equipment and readable storage medium |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107798032A (en) * | 2017-02-17 | 2018-03-13 | 平安科技(深圳)有限公司 | Response message treating method and apparatus in self-assisted voice session |
CN109767765A (en) * | 2019-01-17 | 2019-05-17 | 平安科技(深圳)有限公司 | Talk about art matching process and device, storage medium, computer equipment |
CN110177182A (en) * | 2019-04-15 | 2019-08-27 | 中国平安人寿保险股份有限公司 | Sensitive data processing method, device, computer equipment and storage medium |
CN110347863A (en) * | 2019-06-28 | 2019-10-18 | 腾讯科技(深圳)有限公司 | Talk about art recommended method and device and storage medium |
CN110377731A (en) * | 2019-06-18 | 2019-10-25 | 深圳壹账通智能科技有限公司 | Complain text handling method, device, computer equipment and storage medium |
CN111222896A (en) * | 2020-01-17 | 2020-06-02 | 中信百信银行股份有限公司 | Customer complaint problem processing method, apparatus, device and storage medium |
CN111859958A (en) * | 2020-07-23 | 2020-10-30 | 中国平安财产保险股份有限公司 | High-complaint-risk user identification method, complaint early warning method and related equipment |
-
2021
- 2021-03-31 CN CN202110353074.3A patent/CN112860876A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107798032A (en) * | 2017-02-17 | 2018-03-13 | 平安科技(深圳)有限公司 | Response message treating method and apparatus in self-assisted voice session |
CN109767765A (en) * | 2019-01-17 | 2019-05-17 | 平安科技(深圳)有限公司 | Talk about art matching process and device, storage medium, computer equipment |
CN110177182A (en) * | 2019-04-15 | 2019-08-27 | 中国平安人寿保险股份有限公司 | Sensitive data processing method, device, computer equipment and storage medium |
CN110377731A (en) * | 2019-06-18 | 2019-10-25 | 深圳壹账通智能科技有限公司 | Complain text handling method, device, computer equipment and storage medium |
CN110347863A (en) * | 2019-06-28 | 2019-10-18 | 腾讯科技(深圳)有限公司 | Talk about art recommended method and device and storage medium |
CN111222896A (en) * | 2020-01-17 | 2020-06-02 | 中信百信银行股份有限公司 | Customer complaint problem processing method, apparatus, device and storage medium |
CN111859958A (en) * | 2020-07-23 | 2020-10-30 | 中国平安财产保险股份有限公司 | High-complaint-risk user identification method, complaint early warning method and related equipment |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113449350A (en) * | 2021-06-30 | 2021-09-28 | 中国工商银行股份有限公司 | Management method, device, equipment and medium for USB outgoing sensitive information |
CN113822062A (en) * | 2021-09-22 | 2021-12-21 | 未鲲(上海)科技服务有限公司 | Text data processing method, device and equipment and readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109587360B (en) | Electronic device, method for coping with tactical recommendation, and computer-readable storage medium | |
CN109815314B (en) | Intent recognition method, recognition device and computer readable storage medium | |
US10803253B2 (en) | Method and device for extracting point of interest from natural language sentences | |
US10803252B2 (en) | Method and device for extracting attributes associated with centre of interest from natural language sentences | |
CN115828112B (en) | Fault event response method and device, electronic equipment and storage medium | |
CN112860876A (en) | Session auxiliary processing method and device | |
CN109299227B (en) | Information query method and device based on voice recognition | |
CN109766441B (en) | Text classification method, device and system | |
CN114007131A (en) | Video monitoring method and device and related equipment | |
CN107545505B (en) | Method and system for identifying insurance financing product information | |
CN110598008A (en) | Data quality inspection method and device for recorded data and storage medium | |
US9836693B2 (en) | Using cohorts in a question answering system | |
CN113450075A (en) | Work order processing method and device based on natural language technology | |
CN115033677A (en) | Event processing method, device, equipment and medium based on conversation robot | |
CN113011169B (en) | Method, device, equipment and medium for processing conference summary | |
CN110263135B (en) | Data exchange matching method, device, medium and electronic equipment | |
US20140172874A1 (en) | Intelligent analysis queue construction | |
US20200159824A1 (en) | Dynamic Contextual Response Formulation | |
US11947872B1 (en) | Natural language processing platform for automated event analysis, translation, and transcription verification | |
WO2023272833A1 (en) | Data detection method, apparatus and device and readable storage medium | |
CN113010664B (en) | Data processing method and device and computer equipment | |
CN115221892A (en) | Work order data processing method and device, storage medium and electronic equipment | |
US9244901B1 (en) | Automatic speech tagging system and method thereof | |
CN110502630B (en) | Information processing method and device | |
CN115730104A (en) | Live broadcast room processing method, device, equipment and 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 |