The most popular and comprehensive Open Source ECM platform
Let us know more
Data Algebra: A More Rigorous and Mathematical Approach to Data Analysis
Data Algebra is a technique that some are saying can abstract away much of the grunge and headaches caused by many of the current tools and make managing data, especially Big Data, much easier to do. Data algebra is a mathematical approach for analyzing data sets by apply classical set theory.
Bill Rogers, a senior engineer at IBM, said that “you can do far more sophisticated optimization when you’re using algebraic techniques than you can when you’re just using high-level procedural techniques.”
Charles Silver, CEO of Algebraix Data, a company promoting the use of data algebra, said that “surprisingly, mathematics has played almost no role in software and programming until now – even though it has enabled huge advances in virtually every other form of science and engineering. Data algebra provides a universal language for data and can be the technology that allows the thousands of different data models in the world to be integrated.”
Algebraix Data is promoting data algebra and has open-sourced a Python library that implements the concepts that are explained in a freely downloadable book. The abstract notation of the approach is certainly elegant and it also frames the process of data analysis in a mathematically rigorous way.
But SQL is already a widely-used programming language based on ideas of set theory? What’s different?
The Algebraix Data book calls out SQL as “a mathematical disaster of the first order”, and points out troublesome aspects of the language, like the use of a null value, a concept described as a “wholly unmathematical idea”. Also, because SQL only can handle data in tabular format other types of databases have needed step in where an RDBMS database fails, like for hierarchical data, text and graphs. Programmers have also have long had issues with being able to map data created in object models of programming languages into tables of a database, a frequently-discussed problem called “impedence mismatch”.
The concept certainly has merit, but it’s likely to gain traction only after a serious uphill battle against existing databases and analysis tools.
