Design And Research Of Voice Interaction

Under the environment of the third economic transformation caused by the rapid development of artificial intelligence technology,bank outlets need to adapt to the trend of intelligent upgrading and transformation.At present,there are many problems in the financial self-service terminals of major banks,such as inconsistent standards,low degree of intelligence and difficult operation.As the financial self-service terminal is a facing user equipment in the bank network,the intelligent upgrading of the financial self-service terminal can effectively promote the process of bank reform.Based on the above background,this paper proposes to study the voice interaction mode between human and machine and construct the voice user interface of financial self-service terminal by applying the current relatively complete intelligent voice technology and natural language processing technology.Firstly,the case analysis method is used to study the interaction mode of intelligent voice products on the market.It is found that task-oriented dialogue mode is widely used,accurate and reliable,and conforms to the application scenario of financial self-service terminal.Then,in the process of user research,questionnaire survey and script guidance are used to collect the basic information and feature information of the target user,extract the pain points and dig in-depth the needs of the target user for voice user interface.Then,according to the application scenario of financial self-service terminal and the needs of target users,the product form is analyzed,and the visual-auditory dual-modal user interface composed of graphical user interface and voice user interface is determined.The GUI part produces prototype scheme from layout,elements and interaction process,while the Voice User Interface part produces prototype scheme from example dialogue,skill definition and flow chart.Finally,the program of dual-mode user interface is developed in Python 3.6 development environment,and a semantic understanding algorithm for financial self-service terminal usage scenarios is proposed and implemented.Testers are invited to test and evaluate the usage,and further improvement and optimization suggestions are obtained.Through the evaluation,it is found that the financial self-service terminal integrated with intelligent voice technology improves the efficiency of business management by reducing the time spent by users in finding function buttons,switching pages and browsing information layer by layer.The ability of speech user interface to understand natural language makes users not need to use very standard language to express their intention,which improves the usability of the device.The design and research of voice user interface of financial self-service terminal not only effectively promotes the process of intelligent upgrading of bank outlets,but also provides some reference for the design and development of related intelligent voice products.