Expert system to advise recipes

I did this funny project in 2005 for university. The main feature of this project is that the project was done using bayesian network in order to make decisions and a language based on rules as CLIPS. Therefore, the system is able to make a decision depending if the user prefers more greasy or more calories and so on, being careful not repeat a made decision in last days and if the user has the enough ingredients in the warehouse.

• Firstly, the system provides a screen in order to include new recipes. The recipes will be stored in a database. Some required data must be filled in order to enable to the system to work.

Also, the warehouse must have enough ingredients in order to be able to select recipes with almost all their ingredients. The system will subtract the ingredients of the recipe when this recipe be selected.

• Secondly, the user must specify what she/he prefers. These data will be diffuse values. Therefore, the user will specify if she/he wants a lot of calories, or not much greasy and so on. This is very interesing because these data had been stored in exact data. 

• Finally, the system will value all these data in order to select the best recipe according to the following weights:

- What the user prefers.
- What the user has selected the previous days.
- What the user has in the warehouse.

And the system is able to offer results: