Electronic farming records ? A framework for normalising agronomic knowledge discovery

Nowadays, agriculture generates huge data volumes through different sources. The analysis of this big data enables extraction and deduction of useful knowledge that can be exploited at various levels of the farming process and by different actors (farmers, companies, and agronomists). Nevertheless, to analyse this data effectually, various data sources need to be standardized and integrated into a unified dataset. In this paper, we propose an electronic farming record model using a fact constellation schema that is flexible enough to incorporate various farming datasets and big data models. We also apply some analysis techniques to extract knowledge with the view to optimise crop yield, reduce cost and pesticide resistance, and protect the environment. This can be done by finding suitable quantities of soil properties (texture and pH), soil nutrients, seed rate, herbicides, insecticides, fungicides and adjuvants, which are individually assessed on the most popular crops in Europe. Through experimentation, we show that our models are efficient and very promising.

Automated summary

Agriculture generates large volumes of data that require effective analysis to improve crop production, reduce costs, and protect the environment. This paper presents an electronic farming record model and analysis techniques to optimize crop production and environmental protection.