The human colon is an anaerobic bioreactor with an enormous number of microorganisms, which collectively are named the human colonic microbiota. This community performs the breakdown of polysaccharides that are not digested in the upper intestine, mediating some of the effects of diet on health. As a result of its limited accessibility and its complexity, the human colon remains largely unexplored. Mathematical models can play a central role for a better insight of the human colon complexity. In this context, this work presents the development of a mathematical model of carbohydrate degradation. Our aim was to provide an in silico approach to contribute to a better understanding of the fermentation pattern in such an ecosystem. Our mathematical model is knowledge-based, derived by writing down mass-balance equations. It incorporates physiology of the intestine, metabolic reactions and transport phenomena. As a result, it has a complex structure with a large number of state variables and parameters. In order to handle such a complexity, we made use of mathematical and engineering tools. They included theoretical concepts on identification and model order reduction, and practical aspects of parameter estimation and parameter reduction by knowledge integration. Model simulations provided an adequate qualitative representation of the human colon. The methodology that we have contributed to develop should enable one to assess nutritional scenarios and to address questions that are difficult or impossible to elucidate by experimentation. Our model is a promising tool in predicting the fermentation pattern by human colonic microbiota. It should help in development of strategies aimed at promoting healthy gut function.