Study On The Speech Interaction Of GIS Graphic System

Human-computer interaction is a study of the relationship of human and the hardwareenvironment. The process of interaction is a two-way information exchange between usersand the system, the information includes various of symbols and users’ actions. The essence ofinteraction is a process that the human behavior operates on the system data by using thehardware facilities, to get a certain data operation such as input，converse，process and output,then users get the feedback from the system. With the continuous development of interactiondesign, artificial intelligence, human-computer interaction developed from the traditionalkeyboard and mouse interaction mode to intelligent, multi-channel, multi-media interaction.Speech interaction is one of that intelligent interaction, using human language to expressinteraction semantics.The system can extracts up and recognizes the audio information,translates the audio information. At last, the system’s functions are mapped to thecorresponding response. Speech interaction is known as one of the closest to naturalinteraction.GIS becomes powerful but its interaction remain simple, resulting in inconvenience insome cases. To develop the interaction of GIS, we could not take the method of other systemsmechanically, but have to do more research on the GIS, study its interactive features. Themost important part of GIS is spatial data, which the interaction is operating on. The operationincludes data adding, deleting, modifying and the other processing, some operation deal withthe system environment. In the field of GIS, its interaction semantics are clear, it has a regularinteraction language grammar and rule. This paper main focus on the purpose of introducingspeech interaction to the GIS, the main study are as follows:1. Speech interaction patternsTaking a relatively mature product as an example, we explored the speech interactionmode, summed up that the speech recognition and semantic understanding are the core of thewhole interaction technology.2. Speech recognitionStudying the theory of speech recognition, summing up the model of speech recognition,identify a recognition mode with limited vocabulary, specific person, isolated word. Then using the existing technology, realized a prototype of the speech recognition system, toprepare for the further speech interaction.3. GIS graphic system semantic understandingThe interaction language understanding is limited in the field of GIS graphic system,studying a certain of typical GIS, listing up their common functions, summarizing theinteraction semantics and the features, building GIS interaction vocabulary, separating thewords into nouns and verbs, use concept tree and network structure to complete the modelingof nouns and verbs. After that, we work out a basic language grammar to help understandingthe interaction semantics, it is “noun-verb-noun/adj./adv.”, based on which we make a methodto understand semantics using Schank’s CD theory. Finally, a prototype system is designed totest the feasibility of the speech interaction in GIS graphic system.The main results of the paper are as follows, speech interaction technology can meet thebasic needs of GIS graphics system; The concept knowledge are capable to express thesemantic concept of operations and human behavior. Speech interaction will also be a friendlyinteraction patten in interaction environment such as survey environment in the future.