OLAP, MOLAP and ROLAP: What`s next?

Many believe that, today, information is replacing capital as a key determinant of wealth.

One of the most valuable assets a company owns it its information repository.

In order to gain an insight into the data held in this repository for trends analysis, performance analysis and the formulation of corporate decision-making information, online analytical processing (OLAP) technology was conceived almost two decades ago.

OLAP is a category of software technology that allowed users to delve deeply into their corporate information, not to simply answer "who?" and "what?", but to tackle the tougher questions "why?" and "what if?"

Using OLAP, executives were, for the first time, able to provide reliable answers to questions relating to business projections, production scheduling and capacities, sales forecasting, distribution logistics, budgets and other challenges.

It is important that OLAP - a data querying tool - is not confused with a data mining tool. Data mining systems use neural networks, statistical regressions and fuzzy logic to perform tasks that are not associated with OLAP and its later evolutions.

Not long after OLAP`s creation, a more sophisticated form of the technology was demanded to support calculations and modelling applied across multiple dimensions, through hierarchies and across sequential time periods

Multi-dimensional OLAP (MOLAP) was born. MOLAP helps the user to "slice and dice" information, providing multi-dimensional analyses of data by putting data in a cube structure.

Most successful MOLAP products use a multi-cube approach in which a series of small, dense, pre-calculated cubes make up a hypercube.

For example, MOLAP is the ideal tool to use if you need to know answers to questions that involve a time-line (day, week, month, year), geographical areas (city, state, country), various product lines or categories and channels (sales people, stores, etc).

In other words, the MOLAP approach will allow you to align and analyse data relating to the number of products sold, by sales person to any number of customers in various industry categories within regions in specific time periods.

MOLAP is also able to align data content with the analyst`s own views, helping to eliminate confusion and lower the incidence of erroneous interpretations.

Furthermore, because the data is physically stored in a multi-dimensional hypercube, the speed of these operations became faster and more consistent.

The combination of simplicity and speed are some of the key benefits of MOLAP.

Nevertheless, the efficiency of MOLAP tools is constrained by the limitations surrounding text based query results.

This, and the need for a query mechanism to address multiple, related data tables resulted in the development of relational on-line analytical processing (ROLAP).

The basic difference between MOLAP and ROLAP technologies is an architectural one: MOLAP products take the needed data and put it in a specialised data cache.

ROLAP provides for similar functionality, but it performs its analysis on the fly, without the intermediate step of placing the data in a special server.

This has the advantage of allowing users to quickly make queries against databases that are too large for MOLAP to efficiently parse through.

However, ROLAP does have its limitations. It is cumbersome, difficult to maintain and slow - particularly when multiple tables need to be addressed.

The biggest database of all - the Internet - is likely to transform the role of OLAP technologies in the business environment because of its emphasis on real-time transaction processing.

There is a growing need to deliver real-time information in many areas of commerce and industry. Today, business leaders expect a wide range of information to be available immediately and be accurate up to minute.

In today`s post-relational era, Web and other applications that can create links to objects from any source to meet these real-time goals are aggregating under the hybrid-OLAP (HOLAP) banner.

HOLAP represents, in effect, the convergence of MOLAP and ROLAP. It is a common platform from which MOLAP and ROLAP transactions can be performed very quickly.

It combines traditional data-driven analysis tools with hypertext links to related descriptions and other kinds of information, irrespective of format, making a much richer form of analysis available to the user.

In future, HOLAP solutions will enable users to access data warehouse and decision support applications through standard Internet browsers in real time, gearing up performance, query flexibility, and enhancing the user`s ability to leverage installed relational database management systems.

These Web-based solutions will offer users unparalleled ad hoc functionality and the ability to integrate additional, often diverse, data elements into analyse cubes on the fly. From a practical perspective, HOLAP will allow users to address queries in areas as important and diverse as fraud detection and supply chain management in real-time and on the fly.