The Finish Food Authority (FFA) is using artificial intelligence (AI) technology to dramatically reduce the time it takes to complete critical research as it strives to prevent potential public health crises.
The FFA’s Interdisciplinary Risk Assessment Unit works on scientific risk assessment projects, including the effect of avian influenza on food safety.
Traditionally, researchers would manually read and process tens of thousands of documents while investigating, which would mean projects take two to three years from start to finish.
This is problematic in cases such as controlling the effects where the organisation needs to act quickly. For example, traditional text search could retrieve up to 20,000 pieces of research on a subject, which researchers would manually go through to identify the useful papers. Now using AI from Iris.ai to automate document reading and processing, the FFA can address issues in real time.
The FFA is now using natural language processing expertise from Iris.ai as well as the supplier’s Researcher Workspace platform.
Founded in 2019, The FFA is the result of the merger of the Finnish Food Safety Authority, the Agency for Rural Affairs, as well as part of the IT services of the National Land Survey of Finland. Its main responsibilities are to promote, monitor and study the safety and quality of food, maintain the health of animals and plants, and oversee animal feeds and plant protection products.
“Project delivery times were a critical challenge as the results and insights needed often would be time-sensitive,” said the FFA.
For example, when the FFA was working on a project to understand the factors influencing risks from avian influenza, work “needed to happen rapidly as there was a need to understand the possible spread of the disease on different regions of the country”, said the FFA.
The FFA added: “Manually looking through [papers on avian flu] would have been an exceptionally tedious task given the niche area of the research, which was exploring the intersection of the biosecurity of farms and routes of migratory birds.
“Yet, using Iris.ai’s tools, especially the visual interfaces feature of its Exploration maps, [the FFA] could narrow down the relevant papers across disciplines to that intersection, allowing researchers to find the relevant information. The time saved is particularly important when in crisis situations when issues happen in real-time.”
It is also challenging to find research papers in emerging fields because they are limited, with up to a third of the time used for the project going into information search, according to the FFA.
“Working so expansively across these different fields is challenging. Although FFA researchers may have deep and broad expertise in some matters, there still may be knowledge gaps in others. Therefore, regardless of comprehensive network of researchers and extensive cooperation, every FFA researcher needs a fast route to the sources of researched information,” said the FFA.
The AI technology means faster paper-searching time, overcoming the challenge of drawn-out material collection period. It enables users to ask research questions, without knowing the exact search terms of the research field, and it produces a visual map of relevant papers categorised by topics.
“This is proven to help researchers find more contextually appropriate papers and get a better overview of the fields of research, which helps in drawing conclusions,” according to the FFA.
The FFA said it has “improved the research stage of papers by creating a more time-efficient data collecting process while including the latest papers in the fields. This creates a more valuable way to manage project deliveries and people’s capacity.”