Design Of Agent Applied To Text Mining System

Applied research on agent modeling and design for text mining, a better study subject with practicability and foreseeability, focuses upon design and modeling for mining facade agent, text transformation agent and text mining agent, aiming at constructing a primary application framework in order to provide agent developers with a practical engineering methodology.Agent architecture shows aspects of objectization. At present, various agent platforms implemented by Java well up ever-increasingly. AOP is a special case OOP, and object-like agents compared with the later take on no more than mental states, belief and capability. Agent oriented software modeling is based on patterns, different from object oriented software modeling and design based on attributes and behaviors.Each element of agent abstract architecture is a headstone for R&D on agent applications. 36 abstract elements grouped into 2 types of agent and service are summed up. Aside from encoding service, agent directory service, message transport service and service directory service constitute faithfully pertinent services.PAC architecture conforms to needs and goals of current text mining systems, which consist primarily of UI, data transformation and data mining function modules et al.. They are brought into more abstract agent, effectively meeting the need of intelligent interaction between different modules and come farther to the task planning.Interface agent emphasizes pro-active, learning ability and self-adaptablity. As a human assitant, it can substitute to achieve task decision by collaboration with users. Data transformation is primary. The main characterics of data warehouse depend on retrieving, transformating and loading data from running resources to a data warehouse and data marts. Transformation covers data retrieving, transformation and loading. In fact, it involves transformation whether you need to store, retrieve or present intention presentation whenever transformation from a data warehouse to another is required.Text mining agent is the application of mathematical or statistical processes for the purpose of extracting hidden knowledge from large data sets, the function of completing the recognition of data mode, discovering undiscovered modes and rules, and the task of digging out classification, aggregation and association rules.It is the practical method about mining system agent modeling and design that makes the R&D procedure clearer, more intelligible and more convenient.