New possibilities in Data Analysis with formulas stored in database fields: Simple, powerful and innovative.

Actually, it is strange that you cannot save formulas in your database. Just as name or salary might be relevant properties of an object, formulas might be too. For example, consider the bonus you have agreed on with your employer, the criteria by which you are looking for a new car to buy, the taxes you must pay per year, criteria for segmentation of a population or identification of risk factors. If formulas can be stored in a database, a range of problems become easily solvable. Problems that had otherwise to be solved by writing software outside the database for much more money and time and not suitable for ad hoc data analysis. Colbert has researched methods to save formulas in tables in a database and methods to apply these formulas to data from other tables at a later time. Important topics in this research are the actual definition and storage of formulas, a strong static type system for formulas and scope rules for applying these formulas to data from other tables. The main challenge is to integrate this in a seamless, simple and intuitive manner into SQL without modifying the language. Colbert has succeeded in developing a simple, powerful and innovative concept which makes it possible to use formula fields just like normal fields in SQL. In this talk we discuss the important topics in the research, and the developed concept. Furthermore we look into several areas of application like criteria matching, data segmentation, risk analysis and bonus calculation.