The food safety community is generating a variety of scientific knowledge, e.g. scientific publications, experimental data, databases, mathematical models and software tools for model generation and application. However, the access to this knowledge and the exchange of information between databases and software tools are currently difficult and time consuming. Therefore, three European institutions specialized in food safety risk assessment (ANSES, BfR and DTU Food) initiated a joint project to establish new community resources facilitating the efficient knowledge integration and exchange into and between IT-based applications and resources. The envisaged “Risk Assessment Modelling and Knowledge Integration Platforms” (RAKIP) will be based on harmonized data formats and consistent rules for knowledge annotation. The feasibility of this concept will be exemplified during a one-year, tri-lateral project between ANSES, DTU Food and BfR. The first RAKIP system will encompass data, models and supporting resources generated by the three institutions. As a next step an open access public RAKIP portal will be created, allowing users to access and download reviewed risk assessment models, modules thereof and related data in a harmonized file format. These files can then be imported and executed by software tools supporting the proposed harmonized file format. The envisaged RAKIP portal will facilitate efficient knowledge sharing between interested stakeholders and researchers as existing models can then easily be adapted to specific needs and used as building blocks for future risk assessment models.
Targeted end users
The RAKIP solutions are designed to support the whole risk assessment community including national and international risk assessment agencies, business operators and academic institutions. Modellers, risk assessors and risk managers will benefit most from the RAKIP portal. However other users like research scientists can also make use of the RAKIP resources, e.g. share their experimental data or apply improved modelling tools in the future.
In this sense, the end-users will find in the RAKIP portal a web-based risk assessment model repository that can be freely assessed and searched, and where models can be downloaded, combined and executed. In addition, modellers can contribute their models to the repository by sharing their own models in the harmonized format. It is important to clarify that the RAKIP portal is not primarily intended to be a new “press-the-button” model simulation software. It will be designed such that experts and non-experts can make use of it.
Food Safety Knowledge Metadata
Models and data need to be annotated in a harmonized way. For this, we have defined a structured list of metadata that are relevant for describing data or models in the risk assessment and predictive microbial modelling domains. For each metadata concept it has also been defined if this concept is considered mandatory or not and what cardinality each metadata concept has.
The “generic metadata schema” contains the complete list of metadata concepts that allow to describe in detail all models or data. Appart from the ”generic metadata schema”, we have defined dedicated metadata schema for specific model / data classes, using relevant subset of the generic metadata schema.
The proposed metadata schema can be accessed via:
https://docs.google.com/spreadsheets/d/1R5ofJAIqAywN97cnBY9R9kqj10AvQOs3Gq472n6iDyc/edit?usp=sharing
Community driven Controlled Vocabularies
To support harmonized annotation of food safety knowledge we established an online resource with controlled vocabularies for those metadata concepts that are not free text. Some of them are specific for the different model classes. These controlled vocabularies are based on the terms used by other sources like ontologies, standards and tools (SSD-CODE, FOODON, MIME, PMM-Lab, OpenFSMR, Bibliographic Ontology Specification, etc.)
The lists of proposed controlled vocabularies for food safety knowledge annotation can be accessed via: