Genome-Scale Metabolic Models (GEMs) contain the known chemical reactions in the studied cell or organism. GEMs can be extended to include more biological processes, such as protein translation and secretion. Therefore, GEMs should be integrated with kinetic parameters to couple the metabolic fluxes with the new biological processes. Searching for these kinetics parameters takes a long time. Now millions of scientific papers are available online in the PubMed database. This tremendous amount of current information with text mining approaches creates new opportunities for finding kinetic information that improves GEMs predictions. This paper introduces GEMminer that uses text mining approaches to find this kinetic information based on the user-defined queries and filter keywords. We have validated the accuracy of GEMminer with a manually curated list of GEM papers. Additionally, we have demonstrated that GEMminer can be used to search for kinetic parameters in the protein secretory machinery in yeast, even though there is no available database containing such parameters. These results show that the proposed GEMminer toolbox is a good search tool in PubMed with user-defined queries and filters.