CN110769124B - Electric power marketing customer communication system - Google Patents
Electric power marketing customer communication system Download PDFInfo
- Publication number
- CN110769124B CN110769124B CN201911044565.9A CN201911044565A CN110769124B CN 110769124 B CN110769124 B CN 110769124B CN 201911044565 A CN201911044565 A CN 201911044565A CN 110769124 B CN110769124 B CN 110769124B
- Authority
- CN
- China
- Prior art keywords
- voice
- server
- service
- outbound
- recognition
- 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.)
- Active
Links
- 238000004891 communication Methods 0.000 title claims abstract description 15
- 238000007726 management method Methods 0.000 claims abstract description 31
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 28
- 238000003786 synthesis reaction Methods 0.000 claims abstract description 28
- 238000000034 method Methods 0.000 claims abstract description 26
- 238000005516 engineering process Methods 0.000 claims abstract description 17
- 230000003993 interaction Effects 0.000 claims abstract description 16
- 238000004458 analytical method Methods 0.000 claims abstract description 8
- 238000013518 transcription Methods 0.000 claims abstract description 5
- 230000035897 transcription Effects 0.000 claims abstract description 5
- 230000006870 function Effects 0.000 claims description 29
- 238000012546 transfer Methods 0.000 claims description 26
- 238000012545 processing Methods 0.000 claims description 18
- 238000013475 authorization Methods 0.000 claims description 15
- 230000008569 process Effects 0.000 claims description 12
- 238000012544 monitoring process Methods 0.000 claims description 11
- 230000004044 response Effects 0.000 claims description 11
- 239000012792 core layer Substances 0.000 claims description 9
- 239000010410 layer Substances 0.000 claims description 9
- 238000012423 maintenance Methods 0.000 claims description 9
- 238000013461 design Methods 0.000 claims description 6
- 238000013500 data storage Methods 0.000 claims description 5
- 238000012217 deletion Methods 0.000 claims description 3
- 230000037430 deletion Effects 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000006872 improvement Effects 0.000 claims description 3
- 210000001503 joint Anatomy 0.000 claims description 3
- 238000012986 modification Methods 0.000 claims description 3
- 230000004048 modification Effects 0.000 claims description 3
- 230000002194 synthesizing effect Effects 0.000 claims description 3
- 230000005611 electricity Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 6
- 238000011161 development Methods 0.000 description 4
- 238000007405 data analysis Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000013523 data management Methods 0.000 description 2
- 230000000977 initiatory effect Effects 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000002432 robotic surgery Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/487—Arrangements for providing information services, e.g. recorded voice services or time announcements
- H04M3/493—Interactive information services, e.g. directory enquiries ; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals
- H04M3/4936—Speech interaction details
-
- 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
- G10L13/00—Speech synthesis; Text to speech systems
- G10L13/08—Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination
-
- 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
- G10L15/08—Speech classification or search
- G10L15/18—Speech classification or search using natural language modelling
- G10L15/1822—Parsing for meaning understanding
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/14—Session management
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
- H04M3/51—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
- H04M3/5183—Call or contact centers with computer-telephony arrangements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Networks & Wireless Communication (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Marketing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Telephonic Communication Services (AREA)
Abstract
The invention discloses a power marketing customer communication system which utilizes technologies such as voice recognition, natural language understanding and voice synthesis to respectively realize the transcription of user voice data, the analysis of user intention, knowledge feedback voice synthesis broadcasting and the like, and is matched with an intelligent voice technology library extremely suitable for marketing work characteristics of power grid enterprises to further realize voice interaction between intelligent services and users. The method can well replace the work with high business success rate, and is mainly applied to typical scenes such as customer electric charge reminding and payment urging, data checking, personalized service and the like. Replace 80% of artificial telephones, and achieve the effects of reducing cost, improving efficiency, improving service and benefiting management.
Description
Technical Field
The invention relates to a power marketing customer communication system, and belongs to the technical field of artificial intelligence voice.
Background
At present, national grid company is accelerating to build ubiquitous electric power thing networking, and the development of the integration of "the big cloud thing moves intelligence" technique and marketing business has become inevitable trend, and all kinds of novel marketing services will emerge constantly on this basis, but these final falling to the ground of service need timely acquireing information of user and good service communication, because traditional manpower communication exists with high costs, inefficiency, the big short slab of risk, and the science and technology development has been discomforted. Therefore, the development of the intelligent voice system which accords with the working characteristics and the flow of the power enterprise and the typical customer service scene has very important application significance.
Disclosure of Invention
The invention aims to provide a power marketing customer communication system, and aims to solve the problems that power grid enterprises are wide in customer number and different in demand, and a large amount of marketing service work is too dependent on manpower. The intelligent voice technology is applied to typical scenes such as electric charge reminding and urging, customer data checking, personalized service and the like, various service information under the internet of things and the 'last kilometer' of a user can be quickly communicated, information can be accurately transmitted in real time, and the intelligent voice system can well replace the work of relatively simple processes, repeated operation machinery and massive target customers.
The purpose of the invention is realized by the following technical scheme:
an electric power marketing customer communication system, wherein intelligent voice outbound system combines with IVR system to realize the interactive flow of intelligent voice, outbound questionnaire and response knowledge of the intelligent voice outbound system follow the rules of the existing service system, and carry on the intelligent response according to the response knowledge in the existing service system; the intelligent voice outbound system is combined with the current outbound system, an outbound list is screened by a rule set in the existing outbound system, a user name list conforming to the intelligent outbound flow is pushed to the intelligent voice outbound system for outbound, a service flow which cannot be completed in the intelligent voice outbound system can also be automatically transferred to an artificial seat in the existing outbound system for outbound, and outbound data is recorded;
the intelligent voice outbound system consists of a VUI service layer, a core layer, background operation and maintenance management and data service; the VUI service layer is used for realizing the flow design and the session logic design of intelligent voice interaction and realizing the butt joint with a service system through interface service; the core layer is used for providing voice capability support for intelligent voice interaction services, can realize recognition transcription and semantic understanding analysis of user voice data, and can convert character information in a service knowledge base into voice data for real-time broadcasting; the core layer comprises four parts, namely MRCP transfer service, voice recognition service, semantic understanding service and voice synthesis service; operation and maintenance management and database: the functions of managing an outbound flow and data, monitoring the running state of an engine, displaying a voice data processing result and searching a recorded voice tune are supported; statistics and report analysis are carried out on system operation parameters and various data operation results; supporting the management work of user authority and authorization;
MRCP service, which realizes the interface of intelligent voice navigation and IVR platform, MRCP server receives the user voice data from IVR platform, identifies and understands, and sends the identification result (menu node information) to IVR platform, and the IVR platform carries out relevant service operation; meanwhile, the MRCP server also receives the text transmitted by the IVR platform, and the voice synthesized by the TTS engine is provided to the IVR platform, and the voice is played to the user in the service flow;
the scheduling service mainly comprises: the system comprises an Apache Web server, an Audiokeeper transfer module, a Tomcat Web server and a TTS proxy module;
apache Web server: the Web server is based on Apache, mainly realizes the processing of HTTP query requests by using a PHP technology, realizes forwarding and redirection and has extremely high concurrency and processing capacity; the Server is responsible for forwarding the identification request from the MRCP Server, forwarding the identification result to the understanding module after obtaining the identification result, and finally returning the identification and understanding result to the MRCP Server;
audiokeeper transfer module: the transfer service adopts a mode of TCP direct connection, and mainly realizes the transfer function of voice data:
1) continuously receiving voice data from a client;
2) continuously sending the data to a voice recognition server side to finally obtain a recognition result;
the essence of the Audiokeeper is that a plurality of post requests from the front end are spliced into a data packet, and the data packet is in long connection with the rear end, and the Audiokeeper is simultaneously a transfer module and is connected with the Decoder through a socket; in addition, the database is also operated through a network interface; the interaction aspect of the Audiokeeper is responsible for receiving the voice data uploaded by the Apache/PHP and uploading the recognition result returned by the Decoder to the Apache/PHP; the Audiokeeper transfer module is also responsible for load balancing of the Decoder; in the aspect of exchange with a database, the method mainly stores the received voice stream data and the returned recognition result into the database; in addition, the system is also responsible for the authentication of user authorization;
tomcat Web server:
the understanding service is built on a Tomcat server, an open understanding service entrance is realized by using a Java technology, and an understanding task is completed;
TTS proxy module:
the TTS Proxy is a speech synthesis forwarding module and is responsible for forwarding a TTS request, distributing a task to an idle TTS Combiner, completing a synthesis task and sending a synthesized speech sub-packet back to an upper layer module;
the speech/semantic understanding server cluster includes three classes: a speech recognition Server Decoder, a speech synthesis Server TTSCombiner and a semantic understanding Server Understand Server;
the voice recognition server is responsible for receiving the voice input of the user submitted by the transfer server, decoding and recognizing the voice and finally returning the recognition result to the transfer server;
the voice synthesis server is responsible for synthesizing the text content to be fed back to the client into a sound stream or a sound file, and then playing back the sound stream or the sound file to the client;
the semantic understanding server mainly understands the text content, particularly the content of voice recognition, analyzes named entity information and intention information of a user in the text content and particularly the content of voice recognition to provide the information for a human-computer interaction system (such as an intelligent voice navigation system) to perform subsequent operations;
the speech recognition and semantic understanding server cluster is also closely related to the application service system. The application service system provides the contents to be processed (including linguistic data, grammar and word list of voice recognition, text contents to be synthesized, semantic database contents and the like) for the voice processing server group; in the actual online implementation process, an interface is well defined, seamless connection between the two is ensured, and the functions of voice recognition, synthesis and semantic understanding are ensured to be accurately and fully exerted;
operation and maintenance management and database:
the system can support the management of an outbound flow, the display of a voice data processing result and a recording, tuning and searching function, also support the statistics of system operation parameters and various data operation results, and also support the management work of users and authorities; the database mainly provides functions of authorization management, flow monitoring and voice data storage; in the aspect of authorization management, the database server provides management functions for user profiles, including registration, statistics, modification and deletion operations of the user profiles; the system records all the user authorization numbers of the service so as to be used as the basis for providing service and charging for the user; in the aspect of flow monitoring, a database server records the sending and receiving conditions of a data packet requested by a user each time and the time information of each node requested at this time; according to the information, the flow condition of the online system can be conveniently mastered, and the evaluation and the improvement of the service quality of a manager are facilitated; in the aspect of voice data storage, the database server can store voice data and corresponding recognition results of each conversation of a user, and manual marking is carried out by utilizing the voice data and the recognition results, so that the recognition rate of the voice recognition engine is further improved.
The working method of the intelligent voice outbound system of the electric power marketing customer communication system comprises the following steps:
1. actively calling a user, connecting a call, and playing guidance opening white voice by the system; the playing voice can be pre-recorded data, or mixed voice through TTS synthesis service or the combination of the two;
2. after the call is connected, the call center sends a starting identification command to the MRCP Server, and the MRCP Server starts to detect whether the client starts speaking;
3, MRCP Server detects the voice, starts to recognize, and ends the open field white voice playing at the same time;
the MRCP Server completes the recognition and returns the recognition and understanding result at the same time;
5. the intelligent voice outbound system determines the next process according to the understanding result, plays the corresponding prompt voice to the client and simultaneously returns to the step 2 to wait for the voice instruction of the client;
6. the intelligent voice outbound system work ending condition comprises that a client actively hangs up, changes into manual work, changes into specific service processing and waits overtime.
Compared with the prior art, the invention has the beneficial effects that: the system utilizes the technologies of voice recognition, natural language understanding, voice synthesis and the like to respectively realize the transcription of user voice data, the analysis of user intention, knowledge feedback voice synthesis broadcasting and the like, and is matched with an intelligent voice technology library which is extremely suitable for the marketing work characteristics of power grid enterprises to further realize the voice interaction between intelligent services and users. The method can well replace the work with high business success rate, and is mainly applied to typical scenes such as customer electric charge reminding and payment urging, data checking, personalized service and the like. Replace 80% of artificial telephones, and achieve the effects of reducing cost, improving efficiency, improving service and benefiting management.
Drawings
FIG. 1 is a screenshot of a full-process monitoring interface of the present invention;
FIG. 2 is a screenshot of a label rating interface of the present invention;
FIG. 3 is a screenshot of a data analysis interface of the present invention;
FIG. 4 is an overall architecture diagram of the power marketing customer communication system of the present invention;
FIG. 5 is an intelligent outbound flow diagram of the present invention;
FIG. 6 is a data flow diagram of the present invention;
FIG. 7 is a schematic diagram of the working method of the present invention.
Detailed Description
The invention is further described with reference to the following figures and specific examples.
The invention has the technical characteristics that:
1. intelligent dialogue
The system has the core advantages that the system has a sub-scene intelligent voice technology library and a data management mode with corresponding intellectual property rights, can intelligently respond to users, supports an endpoint detection (VAD) function, can immediately interrupt the current voice playing of the robot, enters the intelligent interruption mode, firstly stops calling a corresponding scene configuration flow and a corresponding voice technology, guides the users to return to the original flow. The response is quick, and the customer experience is good.
2. Full-process monitoring
As shown in fig. 1, a background manager can monitor the task state, and can perform call data query, export, and statistics.
3. Automatic recording
The system telephone recording is stored in the cloud, so that the calling at any time is supported, and the recording can be accurately converted into characters.
4. Precision grading
As shown in fig. 2, the data calculation function can judge the user tags in the whole process, accurately lock the target client, and assist in mining the potential value of the client.
5. Data analysis
As shown in FIG. 3, task data is presented clearly to assist in enterprise query management. And the call data and the client state are mastered at any time.
6. Human-machine collaboration
In the process of the robot call, the on-line switching of the call to manual customer service without hanging up is supported, the dialog recording text is displayed by a screen flicking mode, the keywords are highlighted, and the manual operation can be conveniently and rapidly switched into the dialog.
Perfect combination of TTS synthesized speech and artificial recording
The variable part adopts TTS to synthesize voice, the constant part adopts manual recording, two sound sources are optimized and then synthesized into a section of audio frequency to be played, and the whole voice is smooth.
Fig. 4 is a diagram showing the overall architecture of the power marketing customer communication system.
The intelligent voice outbound system needs to be combined with the IVR system to realize an intelligent voice interaction process, and simultaneously, an outbound questionnaire and response knowledge of the intelligent outbound system need to follow the rules of the existing business system and carry out intelligent response according to the response knowledge in the existing business system. Meanwhile, the intelligent outbound system needs to be combined with the current outbound system for use, an outbound list can be screened through rules set in the existing outbound system, a user name list conforming to the intelligent outbound flow is pushed to the intelligent outbound system for outbound, a service flow which cannot be completed in the intelligent outbound system can be automatically transferred to an artificial seat in the existing outbound system for outbound, and outbound data is recorded.
The intelligent outbound system comprises four parts of a VUI service layer, a core layer, background operation and maintenance management and data service, and the functions and the specific description of each part are as follows:
VUI service layer:
the method is used for realizing the flow design and the conversation logic design of intelligent voice interaction and realizing the butt joint with a business system through interface service.
Core layer:
the voice recognition and translation system is used for providing voice capability support for intelligent voice interaction services, can realize recognition and translation and semantic understanding analysis of user voice data, and can convert text information in a service knowledge base into voice data for real-time broadcasting. The core layer mainly includes four parts, which are MRCP relay service, speech recognition service, semantic understanding service, and speech synthesis service, and the specific introduction of each module is described in the following sections.
Operation and maintenance management and database: the system can support the functions of managing an outbound flow and data, monitoring the running state of an engine, displaying a voice data processing result, listening and searching a recorded sound, and the like, and also support the statistics and report analysis of system running parameters and running results of various data. Management of user rights and authorizations is also supported.
Functional description of each module:
the specific functions of each module are described as follows:
MRCP service
MRCP Server, MRCP service, realizes the interface of intelligent voice navigation and IVR platform, MRCP Server receives the user voice data from IVR platform, identifies and understands, and sends the identification result (menu node information) to IVR platform, and the IVR platform carries out relevant service operation. Meanwhile, the MRCP server also receives the text transmitted by the IVR platform, provides the voice synthesized by the TTS engine to the IVR platform, and plays the voice to the user in the service flow.
Transfer scheduling service
The scheduling service mainly comprises: the system comprises an Apache Web server, an Audiokeeper transfer module, a Tomcat Web server, a TTS proxy module and the like.
Apache Web server:
the Web server is based on Apache, mainly realizes the processing of HTTP query requests by using a PHP technology, realizes forwarding and redirection, and has extremely high concurrency and processing capabilities. The Server is responsible for forwarding the identification request from the MRCP Server, forwarding the identification result to the understanding module after obtaining the identification result, and finally returning the identification and understanding result to the MRCP Server together.
Audiokeeper transfer module:
the transfer service adopts a mode of TCP direct connection, and mainly realizes the transfer function of voice data:
1. voice data from the client is continuously received.
2. And continuously sending the data to a voice recognition server side to finally obtain a recognition result.
The essence of Audiokeeper is to put multiple post requests from the front end together into one packet and maintain a long connection with the back end. The Audiokeeper is also a transfer module and is connected with the Decoder through a socket. In addition, the database is also operated through a network interface. And the interaction aspect of the Audiokeeper is responsible for receiving the voice data uploaded by the Apache/PHP and uploading the recognition result returned by the Decoder to the Apache/PHP.
In addition, the module is also responsible for the load balancing of the Decoder. In the aspect of exchange with the database, the received voice stream data and the returned recognition result are mainly stored in the database. In addition, the system is also responsible for the authentication of user authorization.
Tomcat Web server:
the understanding service is built on a Tomcat server, and an open understanding service entrance is realized and an understanding task is completed by using Java technology.
TTS proxy module:
the TTS Proxy is a speech synthesis forwarding module and is responsible for forwarding a TTS request, distributing a task to an idle TTS Combiner, completing a synthesis task, and sending a synthesized speech sub-packet back to an upper layer module.
Speech recognition/synthesis/understanding service
The speech/semantic understanding server cluster mainly comprises three types: a speech recognition Server Decoder, a speech synthesis Server TTS Combiner and a semantic understanding Server Understand Server.
The voice recognition server is responsible for receiving the voice input of the user submitted by the transit server, decoding and recognizing the voice, and finally returning the recognition result to the transit server. The recognition rate of the system should keep the characteristics of high recognition accuracy and quick response.
The voice synthesis server is responsible for synthesizing the text content required to be fed back to the client into a sound stream or a sound file, and then the sound stream or the sound file can be played back to the client.
The semantic understanding server mainly understands the text content, particularly the content of voice recognition, analyzes named entity information and intention information of a user in the text content, and provides the information for a human-computer interaction system (such as an intelligent voice navigation system) to perform subsequent operations.
It is worth mentioning that the speech recognition and semantic understanding server farm is also closely related to the application business system. The application service system provides the speech processing server group with the contents to be processed (including the linguistic data, grammar and word list of speech recognition, the text contents to be synthesized, the semantic database contents and the like). In the actual online implementation process, an interface is well defined, seamless connection between the two is guaranteed, and functions of voice recognition, synthesis and semantic understanding are guaranteed to be accurately and fully exerted.
Operation and maintenance management and database:
the system can support the functions of management of an outbound flow, display of a voice data processing result, record listening and retrieval and the like, statistics of system operation parameters and various data operation results, and management work of users and authorities. The database mainly provides functions of authorization management, flow monitoring and voice data storage. In the aspect of authorization management, the database server provides management functions for user profiles, including registration, statistics, modification, deletion and the like of the user profiles. The system records all the user authorization numbers of the service, so as to be used as the basis for providing service and charging for the user. The information of the user profile should be as detailed as possible, capable of supporting the needs of various services developed on the platform, and supporting off-site user access through a networked interface.
In the aspect of flow monitoring, the database server records the sending and receiving conditions of the data packet requested by each user and the time information of each node requested at this time. According to the information, the traffic condition of the online system can be conveniently mastered, and the evaluation and the improvement of the service quality of a manager are facilitated. In terms of voice profile storage, the database server may store voice data and corresponding recognition results for each session of the user. By using the voice data and the recognition result, manual marking can be carried out, and the recognition rate of the voice recognition engine is further improved.
The implementation method of the invention comprises the following steps:
through the voice recognition (ASR) and voice synthesis (TTS) technologies of the industry center, the intelligent outbound robot can communicate with the set target client according to the preset dialect. The business target of the outbound is achieved through multi-turn conversation and conversational guidance. The telephone can be actively called out, multiple rounds of voice communication can be carried out with the user, and the telephone can successfully call out in the fields of electric charge reminding and payment prompting, customer data checking, personalized service and the like instead of most manual work. The invention can carry out the electric bill reminding by calling out the defaulting list and the account information of the marketing system through the telephone and combining with the mature reminding call-up operation, generates the interactive result of the person to be reminded and the robot in real time and can follow up with the manual work in real time according to the business condition.
Fig. 5 shows an intelligent outbound flow chart of the present invention, wherein a company outbound displays a list of numbers: and a pre-configured outbound display number pool is provided for the intelligent outbound. Robotic surgery list: the method is a preset power outbound business conversation template. And (3) creating a task: intelligent outbound task creation is carried out through task management, and explicit numbers, outbound calling techniques, outbound robot number, outbound time period control and the like are selected. Importing client information: in order to import the called contact information, such as information of residents who overdue electricity, in the task interface, incomplete tasks can acquire the information of the rest contacts through retrieval. And (3) starting a task: for intelligent outbound task initiation, a pause/delete task may be performed after initiation. And (3) carrying out a task: and after the intelligent outbound task starts, voice transcription is carried out in real time, the call recording and the call record uploading are completed and are stored in a database, and the call result is called back to the service system for real-time display. And (3) completing the task: and after the task is completed, pushing a task completion result including a task clue, task details, conversation details and the like through a callback interface.
FIG. 6 is a data flow diagram of the present invention; the intelligent voice outbound system work flow mainly comprises:
1. actively calling a user, connecting a call, and playing guidance opening white voice by the system; the played voice can be pre-recorded data, or mixed voice through TTS synthesis service or the combination of the two.
2. After the call is connected, the call center sends a start identification command to the MRCP Server, and the MRCP Server starts to detect whether the client starts speaking.
And 3, the MRCP Server detects the voice, starts to recognize and simultaneously ends the playing of the open field white voice.
And 4, the MRCP Server completes the recognition and returns the recognition and understanding results at the same time.
And 5, the AI outbound call determines the next process according to the understanding result, plays corresponding prompt voice to the client and returns to the step 2 to wait for the voice instruction of the client.
6. The intelligent voice outbound call work ending condition comprises client active hang-up, manual transfer, specific service processing transfer, overtime waiting and the like.
The operation is as follows:
the invention is applied to the reminding of the electric bill, is a perfect solution aiming at the traditional manual reminding which wastes time and labor and has lower effect, can efficiently and reasonably remind and urge money for the clients who drag the arrearage, and can flexibly deal with the universal difficulty and the change of the clients.
The robot is mainly responsible for: the work with high session success rate such as first touch, loss of contact checking and the like is mainly applied to the payment reminding of the first touch of a client and the 'loss of contact checking' work intentionally delinquent.
The conversational scenario is as follows:
the system is an intelligent voice system with industry characteristics developed by a marketing front-line staff aiming at a daily typical customer service communication scene after deeply knowing marketing working characteristics and flow of a power grid enterprise, and has the core advantages of a sub-scene intelligent voice technology library with corresponding intellectual property rights and a data management mode. At present, the scene-based intelligent speech technology library is divided into: the system comprises an electric charge reminding function, an electric charge payment prompting function, an information informing function, a data checking function, an individualized service function and a comprehensive energy service popularization function. The core competitiveness of the system lies in that the dialect library is stable and mature, the marketing work scene of a power grid enterprise is extremely met, the applicability rate is high, and the popularization is strong.
TABLE 1 application scenario-Power supply electric charge urging call operation (resident) not overdue
TABLE 2 application scenario-Power supply electric charge urging call operation (resident) overdue
The electric power collection service implementation method is shown in fig. 7:
And 2, acquiring the contact person through the service system to establish an outbound task.
And 4, finishing the call, wherein the finishing can be manual switching, hanging up of the called party, overtime waiting and the like.
And 5, data callback, call result, call recording and call text callback to a service system for statistics and archiving.
In the future, with the help of the high-speed construction of ubiquitous power internet of things, the rapid development of new businesses such as comprehensive energy sources and the like, the system can develop comprehensive energy service expansion such as electric vehicle popularization, photovoltaic industry propaganda, combined cooling, heating and power supply and the like, and services such as power market user policy notification and transaction matching and the like for target customers through the functions of accurate label classification and big data analysis.
In addition to the above embodiments, the present invention may have other embodiments, and any technical solutions formed by equivalent substitutions or equivalent transformations fall within the scope of the claims of the present invention.
Claims (2)
1. A power marketing customer communication system is characterized in that an intelligent voice outbound system is combined with an IVR system to realize an intelligent voice interaction process, an outbound questionnaire and response knowledge of the intelligent voice outbound system follow the rules of the existing service system, and intelligent response is carried out according to the response knowledge in the existing service system; the intelligent voice outbound system is combined with the current outbound system, an outbound list is screened by a rule set in the existing outbound system, a user name list conforming to the intelligent outbound flow is pushed to the intelligent voice outbound system for outbound, a service flow which cannot be completed in the intelligent voice outbound system can also be automatically transferred to an artificial seat in the existing outbound system for outbound, and outbound data is recorded;
the intelligent voice outbound system consists of a VUI service layer, a core layer, background operation and maintenance management and data service; the VUI service layer is used for realizing the flow design and the session logic design of intelligent voice interaction and realizing the butt joint with a service system through interface service; the core layer is used for providing voice capability support for intelligent voice interaction services, can realize recognition transcription and semantic understanding analysis of user voice data, and can convert character information in a service knowledge base into voice data for real-time broadcasting; the core layer comprises four parts, namely MRCP transfer service, voice recognition service, semantic understanding service and voice synthesis service; operation and maintenance management and database: the functions of managing an outbound flow and data, monitoring the running state of an engine, displaying a voice data processing result and searching a recorded voice tune are supported; statistics and report analysis are carried out on system operation parameters and various data operation results; supporting the management work of user authority and authorization;
MRCP service, which realizes the interface between intelligent voice navigation and IVR platform, MRCP server receives the user voice data from IVR platform, identifies and understands, and sends the identified result to IVR platform, and the IVR platform carries out relevant service operation; meanwhile, the MRCP server also receives the text transmitted by the IVR platform, and the voice synthesized by the TTS engine is provided to the IVR platform, and the voice is played to the user in the service flow;
the scheduling service mainly comprises: the system comprises an Apache Web server, an Audiokeeper transfer module, a Tomcat Web server and a TTS proxy module;
apache Web server: the Web server is based on Apache, mainly realizes the processing of HTTP query requests by using a PHP technology, realizes forwarding and redirection and has extremely high concurrency and processing capacity; the Server is responsible for forwarding the identification request from the MRCP Server, forwarding the identification result to the understanding module after obtaining the identification result, and finally returning the identification and understanding result to the MRCP Server;
audiokeeper transfer module: the transfer service adopts a mode of TCP direct connection, and mainly realizes the transfer function of voice data:
1) continuously receiving voice data from a client;
2) continuously sending the data to a voice recognition server side to finally obtain a recognition result;
the Audiokeeper is essentially that a plurality of post requests from a front end are spliced into a data packet, and the data packet is in long connection with a rear end, is a transfer module and is connected with a voice recognition server Decoder through a socket; in addition, the database is also operated through a network interface; the interaction aspect of the Audiokeeper is responsible for receiving the voice data uploaded by the Apache/PHP and uploading the recognition result returned by the Decoder of the voice recognition server to the Apache/PHP; the Audiokeeper transfer module is also responsible for load balancing of the Decoder of the voice recognition server; in the aspect of exchange with a database, the method mainly stores the received voice stream data and the returned recognition result into the database; in addition, the system is also responsible for the authentication of user authorization;
tomcat Web server:
the understanding service is built on a Tomcat server, an open understanding service entrance is realized by using a Java technology, and an understanding task is completed;
TTS proxy module:
TTS Proxy is a speech synthesis forwarding module and is responsible for forwarding TTS requests, distributing tasks to idle TTSCombiners, completing synthesis tasks and sending synthesized speech sub-packets back to an upper layer module;
the speech recognition and semantic understanding server group comprises three types: a speech recognition Server Decode, a speech synthesis Server TTS Combiner and a semantic understanding Server Understand Server;
the voice recognition server is responsible for receiving the voice input of the user submitted by the transfer server, decoding and recognizing the voice and finally returning the recognition result to the transfer server;
the voice synthesis server is responsible for synthesizing the text content to be fed back to the client into a sound stream or a sound file, and then playing back the sound stream or the sound file to the client;
the semantic understanding server mainly understands the text content, particularly the content of voice recognition, and analyzes named entity information and intention information of a user in the text content so as to provide the information for a human-computer interaction system to perform subsequent operation;
the voice recognition and semantic understanding server group is also closely related to an application service system; the application service system provides the content to be processed for the voice processing server group; in the actual online implementation process, an interface is well defined, seamless connection between the two is ensured, and the functions of voice recognition, synthesis and semantic understanding are ensured to be accurately and fully exerted;
operation and maintenance management and database:
the system can support the management of an outbound flow, the display of a voice data processing result and a recording, tuning and searching function, also support the statistics of system operation parameters and various data operation results, and also support the management work of users and authorities; the database mainly provides functions of authorization management, flow monitoring and voice data storage; in the aspect of authorization management, the database server provides management functions for user profiles, including registration, statistics, modification and deletion operations of the user profiles; the system records all the user authorization numbers of the service so as to be used as the basis for providing service and charging for the user; in the aspect of flow monitoring, a database server records the sending and receiving conditions of a data packet requested by a user each time and the time information of each node requested at this time; according to the information, the flow condition of the online system can be conveniently mastered, and the evaluation and the improvement of the service quality of a manager are facilitated; in the aspect of voice data storage, the database server can store voice data and corresponding recognition results of each conversation of a user, and manual marking is carried out by utilizing the voice data and the recognition results, so that the recognition rate of the voice recognition engine is further improved.
2. The electricity marketing customer communication system of claim 1, wherein the intelligent voice outbound system workflow comprises:
1) actively calling a user, connecting a call, and playing guidance opening white voice by the system; the playing voice can be pre-recorded data, or mixed voice through TTS synthesis service or the combination of the two;
2) after the call is connected, the call center sends a starting identification command to the MRCP Server, and the MRCP Server starts to detect whether the client starts speaking;
3) the MRCP Server detects the voice, starts to recognize and ends the playing of the open field white voice at the same time;
4) the MRCP Server completes the recognition and returns the recognition and understanding results at the same time;
5) the intelligent voice outbound system determines the next process according to the understanding result, plays the corresponding prompt voice to the client and simultaneously returns to the step 2 to wait for the voice instruction of the client;
6) the intelligent voice outbound system work ending condition comprises that a client actively hangs up, changes into manual work, changes into specific service processing and waits overtime.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911044565.9A CN110769124B (en) | 2019-10-30 | 2019-10-30 | Electric power marketing customer communication system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911044565.9A CN110769124B (en) | 2019-10-30 | 2019-10-30 | Electric power marketing customer communication system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110769124A CN110769124A (en) | 2020-02-07 |
| CN110769124B true CN110769124B (en) | 2020-11-06 |
Family
ID=69333084
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911044565.9A Active CN110769124B (en) | 2019-10-30 | 2019-10-30 | Electric power marketing customer communication system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110769124B (en) |
Families Citing this family (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111653264A (en) * | 2020-05-29 | 2020-09-11 | 北京思特奇信息技术股份有限公司 | Arrearage payment urging outbound method and system |
| CN111724781B (en) * | 2020-06-12 | 2023-10-20 | Oppo广东移动通信有限公司 | Audio data storage method, device, terminal and storage medium |
| CN111917726B (en) * | 2020-07-01 | 2022-03-15 | 中国建设银行股份有限公司 | Adaptation layer, voice communication system and control method thereof |
| CN111861509A (en) * | 2020-07-10 | 2020-10-30 | 上汽大通汽车有限公司 | An interactive customer service platform |
| CN112040073B (en) * | 2020-07-15 | 2022-03-08 | 国网山东省电力公司信息通信公司 | Intelligent directory assistance server, directory assistance system and directory assistance method |
| CN111899735A (en) * | 2020-07-23 | 2020-11-06 | 湖南三湘银行股份有限公司 | Intelligent bank user interaction system |
| CN111653262B (en) * | 2020-08-06 | 2020-11-17 | 上海荣数信息技术有限公司 | Intelligent voice interaction system and method |
| CN112150694A (en) * | 2020-08-12 | 2020-12-29 | 国网浙江省电力有限公司营销服务中心 | A kind of intelligent voice electricity bill payment system and method |
| CN112185383A (en) * | 2020-09-30 | 2021-01-05 | 深圳供电局有限公司 | Processing method and system for customer service return visit |
| CN112185339A (en) * | 2020-09-30 | 2021-01-05 | 深圳供电局有限公司 | Voice synthesis processing method and system for power supply intelligent client |
| CN112333341A (en) * | 2020-10-27 | 2021-02-05 | 北京聚通达科技股份有限公司 | Intelligent voice robot system |
| CN112232841A (en) * | 2020-11-09 | 2021-01-15 | 国网江苏省电力有限公司南通供电分公司 | Telephone and technology interaction process suitable for automatic prompting of resident electric charge |
| CN112766855A (en) * | 2021-01-25 | 2021-05-07 | 浙江百应科技有限公司 | Method, system, platform and electronic equipment for improving logistics tail end distribution efficiency |
| CN113645364B (en) * | 2021-06-21 | 2023-08-22 | 国网浙江省电力有限公司金华供电公司 | Intelligent voice outbound method for power dispatching |
| CN114242052A (en) * | 2021-12-14 | 2022-03-25 | 山东远联信息科技有限公司 | Method and device for outgoing call by smart phone |
| CN114584655B (en) * | 2022-03-01 | 2024-10-25 | 上海浦东发展银行股份有限公司 | Automatic health detection device of intelligent voice interaction system |
| CN115080713B (en) * | 2022-05-25 | 2024-10-25 | 上海浦东发展银行股份有限公司 | Intelligent voice training platform system and method thereof |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103634393A (en) * | 2013-11-27 | 2014-03-12 | 广州市聚星源科技有限公司 | IVR (interactive voice response) and realization method thereof |
| US8838081B2 (en) * | 2012-03-19 | 2014-09-16 | Nuance Communications, Inc. | Mobile device applications for computer-telephony systems |
| CN105206273A (en) * | 2015-09-06 | 2015-12-30 | 上海智臻智能网络科技股份有限公司 | Voice transmission control method and system |
| CN105491102A (en) * | 2015-11-10 | 2016-04-13 | 大连凤祥科技有限公司 | Intelligent monitoring system based on cloud computing technology and device thereof |
| CN107331383A (en) * | 2017-06-27 | 2017-11-07 | 苏州咖啦魔哆信息技术有限公司 | One kind is based on artificial intelligence telephone outbound system and its implementation |
| CN109040482A (en) * | 2018-08-09 | 2018-12-18 | 武汉优品楚鼎科技有限公司 | The unattended intelligent phone inquiry method, system and device of field of securities |
| CN109327629A (en) * | 2018-11-01 | 2019-02-12 | 深圳市号互联科技有限公司 | A kind of artificial intelligent voice navigation data processing method |
| CN109829729A (en) * | 2019-01-18 | 2019-05-31 | 青牛智胜(深圳)科技有限公司 | A kind of intelligence outgoing call system and method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104320552B (en) * | 2014-11-13 | 2017-07-04 | 国网浙江省电力公司电力科学研究院 | The emergent traffic route filtering method of IVR intelligence systems based on 95598 big traffics |
| AU2018216823B2 (en) * | 2017-01-31 | 2020-10-15 | Genesys Cloud Services, Inc. | System and method for speech-based interaction resolution |
| CN107862572A (en) * | 2017-11-06 | 2018-03-30 | 北京中电普华信息技术有限公司 | A kind of power marketing business application system |
| CN109413286B (en) * | 2018-10-22 | 2021-07-20 | 北京移数通电讯有限公司 | Intelligent customer service voice response system and method |
-
2019
- 2019-10-30 CN CN201911044565.9A patent/CN110769124B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8838081B2 (en) * | 2012-03-19 | 2014-09-16 | Nuance Communications, Inc. | Mobile device applications for computer-telephony systems |
| CN103634393A (en) * | 2013-11-27 | 2014-03-12 | 广州市聚星源科技有限公司 | IVR (interactive voice response) and realization method thereof |
| CN105206273A (en) * | 2015-09-06 | 2015-12-30 | 上海智臻智能网络科技股份有限公司 | Voice transmission control method and system |
| CN105491102A (en) * | 2015-11-10 | 2016-04-13 | 大连凤祥科技有限公司 | Intelligent monitoring system based on cloud computing technology and device thereof |
| CN107331383A (en) * | 2017-06-27 | 2017-11-07 | 苏州咖啦魔哆信息技术有限公司 | One kind is based on artificial intelligence telephone outbound system and its implementation |
| CN109040482A (en) * | 2018-08-09 | 2018-12-18 | 武汉优品楚鼎科技有限公司 | The unattended intelligent phone inquiry method, system and device of field of securities |
| CN109327629A (en) * | 2018-11-01 | 2019-02-12 | 深圳市号互联科技有限公司 | A kind of artificial intelligent voice navigation data processing method |
| CN109829729A (en) * | 2019-01-18 | 2019-05-31 | 青牛智胜(深圳)科技有限公司 | A kind of intelligence outgoing call system and method |
Non-Patent Citations (2)
| Title |
|---|
| 《呼叫中心CRM系统服务端的研究与设计》;陈立文，詹舒波;《设计研究与应用》;20161130;第37卷(第11期);第68-72页 * |
| 《基于CIM模型的电力营销管理信息系统建设及应用》;李瑞锋;《管理世界》;20130831(第434期);第101-103页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110769124A (en) | 2020-02-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110769124B (en) | Electric power marketing customer communication system | |
| CN111192060B (en) | Full-channel self-service response implementation method based on power IT service | |
| CN111488433B (en) | Artificial intelligence interactive system suitable for bank and capable of improving field experience | |
| CN106409283B (en) | Man-machine mixed interaction system and method based on audio | |
| CN105592237B (en) | Method, device and intelligent customer service robot for session switching | |
| CN101076184B (en) | Method and system for realizing automatic reply | |
| US7415537B1 (en) | Conversational portal for providing conversational browsing and multimedia broadcast on demand | |
| KR102163081B1 (en) | Interactive voice bot server and unmanned counsel system | |
| CN107833574A (en) | Method and apparatus for providing voice service | |
| CN116644145B (en) | Session data processing method, device, equipment and storage medium | |
| CN109413286A (en) | A kind of intelligent customer service voice response system and method | |
| US12033629B2 (en) | Systems and methods for automating voice commands | |
| CN103593340A (en) | Natural expression information processing method, natural expression information processing and responding method, equipment and system | |
| CN112885348A (en) | AI-combined intelligent voice electric marketing method | |
| CN113992803A (en) | Intelligent customer service call center system | |
| CN115643341A (en) | Artificial intelligence customer service response system | |
| CN111858874A (en) | Conversation service processing method, device, equipment and computer readable storage medium | |
| CN113962735A (en) | Intelligent marketing system | |
| CN111988476B (en) | Automatic voice cooperative working method of customer service system | |
| CN117424960A (en) | Intelligent voice service method, device, terminal equipment and storage medium | |
| CN113204623A (en) | Question answering method and device | |
| CN111626684B (en) | Intelligent tax processing method, device, terminal and medium | |
| CN116340482A (en) | Multi-skill customer service auxiliary product based on enterprise WeChat combined with NLP engine | |
| CN112837684A (en) | Business processing method and system, business processing device and readable storage medium | |
| CN110213344B (en) | Multi-center remote sign language online translation system and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |