Information technology systems play an important role in agriculture, since they provide information in a quick and efficient way to support strategic decisions. However, the excess of information available may confuse and hinder the access to specific information. An alternative is the adoption of systems that provide automatic recommendations according to the profile of a user community. This work aimed to develop a recommender system based on data mining techniques to recommend content regarding the sugarcane crop. The adopted recommendation model relies on recommendation lists that are association rules between web pages, produced from the data users' browsing. The system was deployed on the portal of Embrapa Information Agency, which is a web system that aims to organize process, store and disseminate agricultural technological information. Among the results achieved with the system, it can be highlighted a knowledge base built with association rules, that describes the behavior of a user community and can indicate the most important links in web pages concerning sugarcane. With the adoption of this recommender system, the main benefits to users are: a) detailed information of production stages, with indications of reference material and different sources of reliable statistics; b) information that favors the efficient production considering technical, environmental, social and economic aspects; c) support to public and private stakeholders in planning and decision making; d) transfer of current knowledge with a friendly language for the sugarcane industry.
bounce rate; data mining; association rules; agricultural technological information
