TTS boost CIC
2002/07/02
As the fund management company is set up, the fund management company will launch a comprehensive competition in fund marketing, customer service, management performance, talents, etc., China Fund will enter the age of group. In such a context, it is a comprehensive customer service center, and all fund management companies have established a comprehensive customer service center and meet customers from a simple manual consultation service. Many information query requirements improve. At the same time, it should also become an important way to understand the customer, tap customers, and analyze customers in time to adjust their business strategies. Therefore, the fund management company customer service center system must have an open, advanced, forward-looking and scalability. The system must have a rapid development and changing business needs, have a mature and stable product software platform, and has a business process automatically generate tool, a layered structured design, which enables seamless connectivity to the background host and different database servers. Ensure the integration of overall integration with the transfer agent, direct sales system, office system, and future customer relationship management system. At the beginning of 2002, Beijing Huo Zhuo Technology Co., Ltd. and Yanhuang Xinxing Network Technology Co., Ltd. jointly launched a customer center system platform dedicated to fund management companies - Customer Interaction Center (CIC). CIC Customer Service Center System Platform will enable customer service center systems to achieve automatic service, manual services, active services, email services, monitor management knowledge base management, and customer information management.
The CIC system is a customer-oriented, people-oriented, can provide customers with personalized, comprehensive services, and after successful operation, can provide services to all customers including the Fund Management Company. The system provides customers with multimedia automatic answer service and service representative services. Customers can use a variety of media to obtain a fund information, and the fund trading procedures and other comprehensive customer services at any time. The system can also use the active exhaled function to cooperate with the business promotion department to conduct active marketing, implement the active telephone return visit and an appointment. Since most of the customer is used to the phone query, speech synthesis and transmission are a key link in this system. TTS (speech synthesis techniques) as one of the important set keys of the system has played a key meaningful voice report. TTS technology is also known as the literal conversion technology, which is an interactive means of communicating with the computer in natural language. It transforms the computer you generated by your computer, or the external input of the external input is to be able to understand, fluent Chinese spoken output technology, belonging to speech synthesis. The speech is combined to generate a technique of artificial speech by mechanical, electronics.
The basic structure of TTS is (1) linguistics: plays an important role in the literary conversion system, the main simulator's understanding process of natural language - text regulation, semantic analysis, grammatical analysis, etc. Computers can fully understand the text of the input and give various pronunciations required for the latter two parts. (2) Rhythm processing: Synthetic speech is planned to scale the audio segment, such as sound, pitch and sound, etc., so that the synthetic voice can correctly express words, sounds more natural. (3) Acoustic processing: According to the requirements of the result of the previous two parts, the speech is output, that is, the synthesis speech. The main function of TTS is to read the words, words, sentences, articles, punctuation, numbers, operators, and English letters you need to pass. The voice output can be changed at any time when reading the sound, and the base frequency, Say speed, words or sentences are paused, and so, etc., etc., allows computer users to directly sense the relevant information of the computer. As one of the key technologies of intelligent computer research and one of the key technologies of human-machine voice communications, speech synthesis has been widely received by the scientific community. Today, with the breakthroughs in speech synthesis technology, its importance to computer development and social life is increasingly high. Computer application software products developed by voice synthesis technology, almost deep into each industry and each aspect. The TTS technology used in the CIC system was studied and developed by the Huang Xinxing and Tsinghua University. The research and development of the Chinese Jindu Speech Calculation Research Center, which condenses the research results of speech technology research in Tsinghua University for more than 20 years. Tsinghua University computer system human-computer voice dialogue research group was established in 1979, in the national 863, seventh, eighth five, nine-five states, the National Natural Science Foundation of China, has been supported with the support of 211,985 211,985, National Natural Science Foundation. A number of international leading research results. Early speech synthesis technology is often concentrated on the study of speech synthesis algorithm itself, the methods and means of research, and there are many similarities between speech coding. The construction of its acoustic model also produces more channel characteristics of the anthropic oral cavity. Among them, the famous resonance synthesis system with Klatt, which produced synthesis systems based on acoustic parameters such as LPC, LSP, and LMA. These methods are used to establish an acoustic model: first recording sound, these sounds cover the pronunciation of people in the process of pronunciation; the acoustic parameters of these sounds are extracted, and integrate into a complete sound. In the pronunciation process, first select the appropriate acoustic parameters from the loop, then according to the rhythmic parameters obtained in the rhythm model, the speech is generated by the synthetic algorithm. Since the 1990s, the algorithm of the waveform splicing (PSOLA) has become more and more extensive applications in the speech synthesis system. At the same time, as the computer operation rate is increasing and popular, the study of artificial intelligence has obtained a rapid development, and the latest research results in the artificial intelligence field have penetrated a large amount of speech research, leading to speech processing technology and speech synthesis. The study produced a breakthrough leap. In the research and development of Yanhuang Xinxing and Tsinghua University in Chinese speech synthesis, it also broke through the artificial neural network, decision tree, hidden Markov model, etc., the use of these methods, thoroughly changed the research of Chinese speech synthesis The focus, making the study of Chinese speech synthesis breakthrough early focus on the study of simple algorithms, and become a systematic study. The overall research and development of voice synthesis, a new step and is in the most advanced ranks of the international.
At present, the common customer service system, its similar announcement, and the news of the dynamic fund are pre-recorded in manually, and the broadcast is broadcast when the customer needs. But this broadcast method has some obvious disadvantages, and most prominent is totally imaginary. After the release of various announcements and dynamic funds that are changed at any time, if you rely on manual recording, you will need a special person to keep in order to record the latest announcements and various changes. The CIC system using TTS technology, the text file is real-time through TTS technology, and the conversion time or even short can even be calculated in seconds. That is, when the latest announcement of the fund company is announced, there may be a customer immediately inquiring this announcement. After the corresponding dial-in program is completed, TTS can technical handle the text of this announcement within a few seconds and time. Transfer to the client in a voice manner. In addition, the TTS technology is used to broadcast relevant information, and many people can avoid mistakes. Very simple, in manual operation, often misreading the phenomenon of misreading, and subjective factors have great influence on the recording effect. If the recorder is recorded, the emotion is not full, the recorded effect is naturally unable to be satisfactory. These conditions will make customers a big impression of the company's overall impression. After using Yanhuang Nova TTS technology, these headache problems naturally solved. Handu and Yanhuang Xinxin jointly introduced CIC system has a powerful interactive mode and computer information platform, where the Yellow New Star mature TTS technology makes the entire CIC system more complete, under its unique smart voice controller, the voice rhythm of text output Smooth, so that the listener (Changsheng users) feel naturally lubricated when listening to the company, indifferent and oysters without machine speech output. TTS voice synthesis Technical voice library covers national standards, secondary Chinese characters, English interface, automatic identification, English, support, English mix. All sounds are pronounced by real Mandarin as a standard pronunciation, achieving 120--150 Chinese characters / second fast speech synthesis, reading speed up to 3-4 Chinese characters / second, users can hear clear and pleasant sound quality and co-fluid tone. The key technologies involved in the Yellow New Star TTS (ie SionSonic) include: the design, corpus, corpus and labeling tools suitable for Chinese speech synthesis, based on statistical methods for the design of Chinese text processing modules, based on neural network models and statistical models Rhythm model design, speech primary selection algorithm and its training algorithm, the judgment of speech synthesis acoustic processing, the actual situation of system working framework. The features of the SionSonic system are: using the most advanced data drive technology (can be trained), realization of immediate transmission, and handle multiple call requests, the application field can be expanded, the system performance is stable and reliable, can be used under different system platforms The user interface is flexible. As a technique for application, SionSonic system can be derived from different products, including: speech synthesis development kits SDK V3.0; speech synthesis system; speech synthesis server; intelligent voice controller; offline synthetic system; virtual attendant; Xinhua Angler and so on.
A Call Center system with a World Leading Technology can become an important measure for the Fund Company to meet the challenge. Hu Zhuo and Yanhuang Xinxin jointly launched the CIC system to enable the fund company to improve service quality and attitude, closely related to customer relationships, making it the most important window for Fund Company and customer communication.
