Artificial intelligence applied in the analysis of cerrado medicinal plants

• Grant number: 08/10499-4
• Support Opportunities: Regular Research Grants
• Start date: August 01, 2009
• End date: July 31, 2011
• Field of knowledge: Biological Sciences - Genetics - Plant Genetics
• Principal Investigator: Silvana Giuliatti
• Grantee: Silvana Giuliatti
• Host Institution: Faculdade de Medicina de Ribeirão Preto (FMRP). Universidade de São Paulo (USP). Ribeirão Preto , SP, Brazil

Abstract

The Cerrado biome is the largest savanna in the world. Its understanding, in terms of conservation and development, depends on studies that analysis the relationship between many features of the environment and species, with a high degree of complexity. The Bioinformatics has assisted in studies of complex systems through the use of Artificial Intelligence, which allow the crossing of features for analysis and classification of certain states. The objective of this project is the development of an intelligent system applied in the selection and the analysis of characteristics of the environment and the medicinal plants of the savannah that lead them to a specific state. This will use techniques of learning machines to analyze the degree of importance and relationship of these features for classifying the state. The system will be developed for the web, allowing access for different groups of researchers. Initially, the study will address two species: Stryphnodendron astringent and Palicourea rigid. The attributes are derived from data collected in field and laboratory tests. The system will provide a structured environment in which the information obtained on the plants will be registered and categorized by the researcher. By using this data, the algorithms of learning machines can be trained to select relevant characteristics of medicinal plants of the cerrado. The system will assist in the search for answers to many questions, such as, what climate and soil characteristics are important in the development of the plant? Therefore, such issues could have the support of the system in decision-making. (AU)