COTTON PEST-MANAGEMENT - A KNOWLEDGE-BASED SYSTEM TO HANDLE INFORMATION INPUT OVERLOAD

Citation
Mw. Geyer et al., COTTON PEST-MANAGEMENT - A KNOWLEDGE-BASED SYSTEM TO HANDLE INFORMATION INPUT OVERLOAD, AI applications, 8(2), 1994, pp. 1-20
Citations number
NO
Categorie Soggetti
Environmental Sciences","Computer Science Artificial Intelligence",Forestry,Agriculture
Journal title
ISSN journal
10518266
Volume
8
Issue
2
Year of publication
1994
Pages
1 - 20
Database
ISI
SICI code
1051-8266(1994)8:2<1:CP-AKS>2.0.ZU;2-W
Abstract
Successful decisions largely depend on correct interpretation of data. Today, our ability to collect data outstrips our ability to interpret it, a situation called ''information input overload.'' Information in put overload is known to have a deleterious effect on decision makers. Full use of data, knowledge, and other information requires a system that can extract the critical decision factors and follow a decision t ree to find related pieces of information. A knowledge-based system wa s built to aid the project management team responsible for identifying cotton fields at risk to pink bollworm and releasing sterile pink bol lworm to help control the native pink bollworm population. The system uses object-oriented design, expert system techniques, a link to simul ation models, and database management in an integrated system to optim ize, improve, and ease the decision-making process. The system made si gnificantly fewer mistakes than did human decision makers, while assig ning treatments to high and low risk areas. In addition, the system th oroughly documents the decisionmaking process and the resulting recomm endations, thus allowing use of adjuncts such as a GIS and simulation models of pest and crop populations.