There's little question that the relational database model is the most popular and will be for some time. Relational databases include Oracle's database, Microsoft's SQL server, MySQL and Postgres.
Simply look at the database market. It largely consists of Oracle, Microsoft, IBM and Postgres. These companies sell relational databases. It is my opinion that relational will be the dominant model for some time.
I agree with Ian. Some additional evidence (probably a consequence of the market dominance by the big players) is that most NoSQL systems after a while start offering SQL related interfaces like JDBC
I also agree with Ian, relational model DB is the most popular in business, followed by dimensional model (with most of the time is implemented over relational model).
Well, I am sure you already hear about everything depends in your requirements or in your design. The market right now have some preferences for relational models, first of all, the relational database is simple to understand and is easy to learn. But is not the only way, some people prefers OLAP models, very fast but kind of complex to understand.
I think is easier if you explain us your requirements and then we can see which model is best for you.
Historically (the last 30 years) the relational database model under a third normal design has been used by almost all businesses and is reflected in the major software offerings from Oracle, Sql Server and so on.
The issue is that the relational model has issues scaling to Big Data types of data, here non-SQL databases such as Hadoop are being used to deal with very large data.
So, if you have a small amount of complex data then the relational model still works best, if you have large amounts of mostly unstructured data then the newer No SQL databases probably represent the future.
There's a bit of a lift to this, but you'll be rewarded with a very useful skill-set portable across many disciplines.
There are good tools available to model data, including tools from Oracle for MySQL, specifically, MySQL Workbench. Before you go there, you'll need a background in entity-relationship modeling and there are some good books for learning those methods. Richard Barker's (see the link) is still a very good one.
Here's the key: these are meta-techniques. The examples used in the book, which are for an airline, are general and can be ported to many different disciplines. Modeling data is a function of the relationships of one data entity to another. As such it doesn't matter whether the data is from biology, bowling, or business. It's about the ordinal relations of the entities themselves - one-to-one, one-to-many, many-to-many, and so on. That's what you learn when you learn to model data. Do it correctly and, on the assumption you've gathered the right information to begin with - that you've targeted the proper entities and sampled them at the resolution you need - then there are no questions you can imagine asking that you won't be able to get an answer to once you've built the right constraints and business rules into the database.
There are several database engines that allow business support systems in any field. The variables used to select one or the other is the type of use ; transactional or not , security levels , use internal network or web , licensing and support costs . In my experience for complex systems , transactional and would recommend you look ORACLE SQLSERVER and for smaller or web query systems see MYSQL and POSTGRES.
At conceptual level, Snow Flake Model is most successful for handling business applications involving big data.
The data models which are document or key-value based are popular as far as scalability, availability and fast access is concerned. these models are successful in handling unstructured data.
SQL based RDBMS looks obsolete, but is not out of market yet because of consistency it offers. Future data models which are based on structured, semi structured and unstructured data will be most successful, but there is no single model that offers all of CAP properties till date.