Printers based on natural language HCI Research and implementation
|School||Dalian University of Technology|
|Course||Electronics and Communication Engineering|
|Keywords||Human-Computer Interaction Feature information extraction Ontology mapping Natural Language Processing Function body|
In the rapid development of the information society, a wide variety of electronic devices frequently used in people's daily life . However, along with the functional complexity of electronic devices , equipment, complexity of operations will also rise, so a friendly , intelligent devices is critical to users interactively . After many researchers efforts to touch interactive , voice interaction represented many novel interaction techniques continue to emerge. However, these still exist, such as interactive touch interactive interface constraints, bottlenecks in the development of interactive voice and other issues . This simple natural language interaction has become a research hotspot . This thesis is based on printer field , the design of a natural language allows users to control electronic devices interaction methods . Users only need to enter the function you want to achieve a natural language description , the system will automatically extract the device functions and parameters , and thus controls the printer to perform operations. The design of interactive systems as a whole can be divided into three stages: the background knowledge base building phase , functional mapping phase extraction and functional implementation phase. Knowledge Base using ontology (Ontology) to be built , which defines almost all areas of the printer functional concept , functional implementation of the relevant concepts , which can be developed through the body 's internal relationships to support the follow-up conducted in two phases ; description of functionality for users extraction, paper use a \special structure , the kinds of structures generated by extraction function mapping mechanism to ensure the successful execution . Paper in the printer field to build the prototype system , wherein the body defines more than 1000 concept . According to a large number of natural language input test , test accuracy rate can reach 89%. The results show that the system can achieve the basic functions of the printer manufacturers to achieve the desired objectives.