Font Size: a A A

Telecom Business Software Design Based On Natural Language Pocessing And Voice Recognition Method

Posted on:2015-10-12Degree:MasterType:Thesis
Country:ChinaCandidate:L ZhangFull Text:PDF
GTID:2308330473451641Subject:Software engineering
Abstract/Summary:PDF Full Text Request
It is always an important subject of Artificial Intelligence research to recog nize and process natural language. Natural language processing and the other ar eas of artificial intelligence such as Theorem Proving, Problem Solving, Pattern R ecognition, Machine Adversarial, Robotics Science, etc have consistency, of which t he essence is both of them are about information expression and utilization. In an other word, both of them collect various sources of unexhausted information an d output information in a rational methodology through computers.In a communication system, through natural language and speech recognitio n technology, traditional communication service has been transformed into a “ser vice provider” and “life partner”. With language processing and voice communic ation network technology, people can easily query and extract information from the system. Natural language processing and speech recognition technology is gr adually becoming the key technology of human-computer interface in informatio n technology. Voice recognition and synthesis technologies combining enable peo ple input commands by voice instead of keyboard.This essay mainly discusses how to reduce the threshold of the application in intelligent technology in the telecom business and improve the advanced technology reuse through the platform construction of voice recognition and semantic understanding ability. Firstly, the discussion begins from the fundamental significance of artificial intelligence technology application whose representative is SIRI. And it is followed by the analysis of original semantic platform for optimization and ability in application of business calls, data mining, understanding, which can solve the telecom business oriented accumulated corpus problems and improve the accuracy of voice recognition & semantic understanding. By testing the platform prototype, the recall ration and precision can reach a balance, so the telecom can complete the technology accumulation and save the cost of new technology development applied in the telecommunication business.In the last, base on the test and analysis of the performance of semantic p latform, and by setting different simulation scenarios, assessing the indicators su ch as process performance, responding time and data reliability under multi-users’ concurrent. And the conclusion is positive to have an effective and feasible d esign scheme.
Keywords/Search Tags:Artificial Intelligence, Recall, Precision, Semantic Platform
PDF Full Text Request
Related items