The design of text mining-based services adapted to specific needs and usable by non-specialists is efficiently addressed by platforms such as OpenMinTeD. The extraction and formalization of high-quality information possibly requires the use of dozens of specialized tools and resources. OpenMinTeD solution lies in their reuse, adaptation and combination. OpenMinTeD pools these resources and offers all the technology and services needed to design, execute and maintain the processing workflows. The case of food quality presented here is a shining example of how the use of semantic analysis methods applied to a very large collection of documents can contribute to the field of microbiology of food processing and preservation. The formalization of the knowledge extracted from the texts for use in conjunction with other experimental and analytical data is a key point of the approach for researchers and industrial users.