Text Mining System for Non-Expert Miners
This is an incremental application of existing methods to a specific domain (oceanography) for non-expert users.
The paper tackles the problem of extracting knowledge from ocean sensor data by designing a text mining system integrated with service-oriented architecture for the Ocean Information Data System, resulting in a structured approach that segments and analyzes data based on region using pre-processing and mining techniques.
Service oriented architecture integrated with text mining allows services to extract information in a well defined manner. In this paper, it is proposed to design a knowledge extracting system for the Ocean Information Data System. Deployed ARGO floating sensors of INCOIS (Indian National Council for Ocean Information Systems) organization reflects the characteristics of ocean. This is forwarded to the OIDS (Ocean Information Data System). For the data received from OIDS, pre-processing techniques are applied. Pre-processing involves the header retrieval and data separation. Header information is used to identify the region of sensor, whereas data is used in the analysis process of Ocean Information System. Analyzed data is segmented based on the region, by the header value. Mining technique and composition principle is applied on the segments for further analysis. Index Terms-- Service oriented architecture; Text Mining; ARGO floating sensor; INCOIS; OIDS; Pre-processing.