© 2024 SURF
Within the food industry there is a need to be able to rapidly react to changing regulatory requirements and consumer preferences by adjusting recipes, processes, and products. A good knowledge of the properties of food ingredients is crucial in this process. Currently this knowledge is available in scattered heterogeneous resources such as scientific peer-reviewed articles, databases, recipes, food blogs as well as in the experience of food-experts. This prevents, in practice, the efficient integration and use of this knowledge, leading to inefficiency and missed opportunities.
In this project we will build a structured database of properties of food ingredients, focusing in particular on the taste and texture properties. By large-scale collection and text mining on a large number of textual resources, a comprehensive data set on ingredient properties will be created, along with knowledge on the relationships between these ingredients. This database will then be used for to find new potential applications for healthy and taste enhancing ingredient combinations by network-based discovery methods and artificial intelligence algorithms will be used. A concrete focus will be on application questions formulated by the industrial partners. The resulting hypothesis will be validated in a real life setting at the premises of the industrial partners.
The deliverables of this project will be:
- A reusable open-access ingredient database that is accessible via a user-friendly web portal
- A set of state-of-the-art mining algorithms that can address a wide variety of industry driven use cases
- Novel product formulations that can be further developed for the consumer and business2business market