Automation of agricultural consulting processes

Task: Ensuring online, automated consulting processes based on well-organized (monetary and natural - cf. demo), online data assets about farm-clusters of the FADN and about single (anonym) farms and after requesting facts and estimations (incl. checking plausibility) of each concerned objects (regions/enterprises), in the following areas:

  • Reporting: automated interpretation of simple charts/tables and the following similarity analyses (basing on expert system approach)
  • similarity analysis (A): comparison to 'best of' groups (benchmarking)
  • similarity analysis (B): indicator-based interpretations (Y0-models)
  • similarity analysis (C): suspicion generation (under- or overestimations, losing of equilibrium)
  • similarity analysis (D): derivation of production functions
  • similarity analysis (E): forecasting

Special (consistency) modules:

  • Integration of results of individual consulting modules must be evaluated for each object (e.g. region/farm) by a consistency maximizing closing module. For example, generation of final farm-value in the form of spreading of 1000 points basing on the individual classification of those indicator-based Y0-model results that also fluctuate around 1000 points.
  • Special analytical layer: revelation of simulation risk, e.g. there is a possibility that a new Y0 model can reveal status-combinations that are less valid because of the unbalanced input rates, based on another model's statuses that fits into the top 25 %. For example, primary data driven models may be classified as distorted (e.g. unreasonably large stair-distances due to small changes of input). In that case, models may be made to reveal differences in primary data that can be compared to the primary data driven models.

Keywords / areas:

  • Case-specific analysis
  • Online expert systems to explain the models
  • Plausibility and consistency checks
  • Handling of inconsistency of the results
  • Revelation of the connections between indicators/indices
  • Online benchmarking
  • Online and offline software development, data mining, animations that demonstrate development steps and , and geographic information system
  • Preparing results to enter the market
  • Process-automation
  • Template development: on context-free, profession-specific, and indicator-specific levels
  • Development of virtual consultant robots

Co-operation partners (in alphabetical order): (Previous projects, separate curricular tasks, theses, practical courser, PhD-dissertations, integration of innovational and research projects)


NameInstitutionReferencesTasks
Horváth HenriettaSZIE GTK ISZAMTDKimplementation, testing
Kovács LászlóSZIE GTK ISZAMOLAPimplementation
Márta AttilaRIIRCORE/GIKGIKtesting, marketing (HU)
Palatinus MiklósSZIE GTK ISZAMTDKimplementation, testing
Pető IstvánSZIE GSZDIPhD-disszertáció előkészületbentesting, system documentation
Pitlik LászlóMY-XMIAU, ill. OTKAplaning, implementing, testing, marketing (HU)
Prof.Dr.Dr.h.c. Miklós Géza Zilahi-SzabóInstitut für Informatik Justus-Liebig-Universität Giessenzilahi.de ill. emendio.deplaning, testing, marketing (DE)
Sápi AndrásSZIE GTK ISZAMTDKimplementation, testing
Varga ViktorSZIE GTK ISZAMTDKimplementation, testing
Vrabély BalázsSZIE GTK ISZAMOLAPimplementing

Attached documents: (URL)

((Back))