Development of an Information Integration and Knowledge Fusion Platform for Spatial and Time Based Advisory Services: Precision Farming as a Case Study

One of the most challenging knowledge services is to provide information relevant enough to support making effective decisions in real time. Even though many sources of relevant data and knowledge are available on websites at any given time, they are scattered and offer little or no information on the semantic relationships, thus making such sources hard to exploit. This paper proposes an approach to developing a spatial and time-based advisory system by using ontology for aggregating data from heterogeneous databases, and from devices such as climate sensors and mobile phones, using shallow parsing to extract the domain-specific concepts and their attributes from semi-structured text, and using production rules to activate functional knowledge formalized from natural language text that is dispersed across the Web. Precision farming for rice is used as a case study since it relies upon intensive sensing of environmental conditions of the crop, extensive data handling and processing, and farmer knowledge. This work aims to support resource-poor farmers toward higher productivity while minimizing costs. The service offered is to therefore provide personal assistance, thus enhancing a farmer's ability to apply actions effectively according to the crop calendar, i.e. the optimal use of pesticides and nutrients in heterogeneous field situations that affect crop quality and reduce risk.