The DIKW pyramid is a scheme that represents the cognitive process and learning, which puts in a hierarchical relationship four elements: data, information, knowledge, wisdom.This model depicts the process of knowledge as a pyramid consisting of a broad base (raw data) that, following a process of aggregation-contextualization (information) and application-experimentation (knowledge) allows to reach wisdom.
Trying to bring this process into the business environment it is possible to understand how the data held by companies can be managed to obtain information from which to derive knowledge and then allow decision-makers to make correct choices supported by data. In recent years, companies have understood the importance of taking business decisions not only based on the instinct and experience of successful entrepreneurs but to do so in a scientific way, based on the dataset available to the company and third-party sources that can offer an increasingly accurate picture of the scenario in which it operates.
Every company today strives to become data-driven to optimize its choices and gain a competitive advantage. To become a company that truly leverages data to make winning decisions, both strategically and operationally, it is necessary to start with the correct acquisition, management and storage of raw data.
DIKW pyramid: raw data
Raw data is data that has not yet been processed. Signals not yet interpreted, the result of a massive and indiscriminate collection, i.e. symbols that have not yet been associated with a meaning. The value of this raw data depends on the use that the company will be able to make of it, but it is fundamental because the path to information, knowledge and wisdom (which is equivalent to making correct business decisions) starts from this very base, so the company must be able to track and collect it.
As long as data is simply collected, but you don’t have the tools to analyze it or get information from it, it is irrelevant to the company and does not contribute to generating value and remains just numbers and strings of characters that take up space in a DB.
DIKW pyramid: the use of Business Intelligence
Business Intelligence system as a Data Warehouse and a series of ETL (Extract/Transform/Load) processes, i.e. the collection of data from an unlimited number of sources and their subsequent organization and centralization in a single repository.
These tools are preparatory to the evolution of the data and the process that leads the company to extract useful information for the business. A piece of data itself has no intrinsic value until it is contextualized. It is only when it is compared with others and evaluated in a wider context that information can be deduced. At this level of processing, specific and pointed questions such as “what?”, “where?”, “when?” can be answered, giving initial meaning to the data collected. The work of the consultant or Business Intelligence software consists precisely in aggregating and contextualizing data to transform them into information. The more refined the intent of the analysis, the more complex the data models and ETL flows will have to be. To get to grips with these in a more informed way, however, there is a need to advance at least one more step up the DIKW pyramid.
DIKW pyramid: from information to knowledge
The output of business intelligence work is represented by reports that aggregate huge amounts of data by returning tables that link different information, or synthetic charts and dashboards. From these aggregations and representations, decision-makers are concerned with understanding what these data represent and from these, they make choices of actions. Even though the use of software for data acquisition and analysis has grown in recent months, even in the Italian scene, there are still few companies that reach this step of the pyramid because to do so, it is necessary to incorporate data and its management into daily business processes.
DIKW pyramid: from data to wisdom
After gathering all the indications based on the processing of historical data, one can move on to the planning of future actions. The phase of wisdom lies precisely in the application of the acquired knowledge and thus being able to act in the best way for the company. The data at this point become the scaffolding on which the strategy is based and at the same time a yardstick for future evaluation of the progress of the project. The focus of this phase is on future actions and how to perform forecasting analysis and assess the future impact of certain decisions.
DIKW pyramid: from historical data to predictive analysis
The biggest challenge of Data Analysis in recent years is to be able to provide tools capable of performing predictive analysis. Predictive analytics involves the use of a variety of statistical techniques such as predictive modelling, machine learning and data mining to analyze historical and real-time data to provide predictions or unknown events. Predictive models search for patterns in historical and transactional data to identify risks and opportunities, find relationships between variables that enable assessments of risk or risk potentially associated with a particular set of conditions, guiding decision making.
To do this, machine learning algorithms are used, i.e. algorithms that allow the program to “learn” based on data stored in the past and on the correctness of the output generated to be able to predict future trends. This software, such as those developed by Premoneo, are based on mathematical-statistical models and artificial intelligence algorithms that allow estimating future trends and forecasts with a minimum margin of error (usually around 95%), offering to the management decisive support to take correct and more profitable decisions.