The AGRICORE project proposes a novel tool for improving the current capacity to model policies dealing with agriculture by taking advantage of the latest progress in modelling approaches and ICT. Specifically, the AGRICORE tool will be built as an agent-based approach where each farm is to be modelled as an autonomous decision-making entity which individually assesses its own context and makes decisions on the basis of its current situation and expectations. This modelling approach will allow simulating the interaction between farms and their context (which will account for environment, rural integration, ecosystem services, land use and markets) at various geographic scales – from regional to global. To do so, advances in big data, artificial intelligence algorithms, mathematical solvers and cloud computing services will be applied to optimise the extremely-long parameterisation and calibration phase required by current agent-based tools, to better mimic the modelling of farmers’ behaviour and interactions, to credibly assess the local effects of global events and EU policies, and in general to improve policy design, impact assessments and monitoring. The AGRICORE tool will be made as a highly modular and customisable suite, and it will be released as an open-source project so institutions can transparently update and improve the tool as needs arise. The further information about the project can be found in the project webpage (https://agricore-project.eu/partners/).
The aim of the project is to develop an "agent-based, spatial" empirical (quantitative) modeling platform that takes into account socio-economic, environmental and climatic characteristics and risk behaviors of farm enterprises in the European Union, and using this platform to develop agricultural and rural development policy analysis. The inadequacy of the existing general equilibrium and partial equilibrium based models in revealing the impact analysis of the rural development policy practices of the EU is the mainstay of the emergence of this project.
The prominent feature of the project is that it will use digital technologies such as machine learning, big data systematics and classification (ontology and semantics) and artificial intelligence, as well as the use of "agent-based" modeling and spatial micro data.
The main contribution of the Akdeniz University team in the project will be to develop the agricultural land, product market and production factors/inputs modules of this modeling platform. The team is also involved in all work packages of the project and undertaken partial responsibilities. The modeling platform created after the completion of the project will be available to researchers in Turkey. Akdeniz University project team will be able to use the platform free of charge and provide consultancy services to users.
Researchers working in the project team from Akdeniz University will develop their expertise and capacity in the following areas.
- -EU agriculture and rural development policies (monitoring and evaluation indicators, impact analysis models),
- -“agent-based” modelling, machine learning,
- - use of big data (characterizing data using ontology; coding and classification, big data storage),
- -modeling of agricultural land market, agricultural input and output markets.
Keywords: Agent-based impact assessment of agricultural policies, rural development policy analysis tools, Machine Learning, Big Data and Artificial Intelligent Use in Agricultural and Rural Policy Impact Assessment
