1.0 era DaaS (database, data warehouse system) Data as a Service
Engineers Support Decisions
Data Analysis 1.0 → Engineers support decision-making
This was the rise of data warehousing, where customers (businesses) and production processes (transactions) were aggregated into gigantic repositories.
Real progress has been made in the objective understanding of business phenomena, through database engineers managing data, processing data, analyzing data,
In this way, managers can make decisions based on understanding based on the results of data , rather than just intuition .
It has been half a century of development from network/hierarchical databases to relational databases . In the period of lack of data analysis tools, data materials are in the hands of database administrators, and engineers classify, organize, code, store, retrieve and analyze data. maintain. However, the limitation of this stage is that the data is only used within the company, that is, business intelligence activities can only deal with what happened in the past, and cannot predict future trends.
2.0 era IaaS (BI + data analysis tools) Information as a Service
Analysts Support Decisions
Data Analytics 2.0 → Analysts Support Decisions
The limitations of data analysis in the previous stage have become more prominent as major enterprises have stepped out of the comfort zone of data warehouses and tried to use broader methods to conduct more complex analysis.
Enterprises have begun to obtain information through external sources, such as click streams, social media, the Internet, etc. At the same time, the need for new tools has become more and more obvious.
As a result, there is a large demand for professional analysts who can use data analysis tools. Enterprises hope to obtain personnel who can analyze data from a business perspective.
Through business intelligence, data analysts are liberated from a large number of simple data charts analyzed through databases , and enterprise data is integrated through data analysis tools , and business intelligence (BI, Business Intelligence) reporting tools are used to achieve beautiful, clear, modular, Dynamically updated data visualization display,
Enable management or decision makers to make decisions based on the results of analytical tools .
At this stage, the company hopes that employees can help process and analyze large amounts of data through fast processing engines. And, go a step further by visualizing the results of descriptive and diagnostic analytics to detect trends, clusters, and anomalies, and predict future trends .
Enterprises have also opened up competition for data analysis. Companies are not only improving traditional methods such as internal decision-making, but also continuously developing more valuable products and services.
The company is developing at an unimaginable speed and has set up more R&D departments internally, such as a data analysis team composed of data scientists, data engineers, solution architects, and chief analysts.
AaaS (Machine Learning) Answers as a Service in the 3.0 Era
Black Box Support Decision
Data Analysis 3.0 → Black Box Model Supports Decision-Making
Pioneering big data companies are starting to invest in data analytics that underpin customer-facing products, services, and capabilities. They lure users to their sites with better search algorithms, buying suggestions, and targeted advertising, all powered by machine models . The phenomenon of big data has spread rapidly, and today it is not just technology companies that are developing products and services from machine models , but companies in almost every industry.
Machine learning training models are designed to gain knowledge , analyze insights , and predict trends from various forms of data using scientific methods, exploratory processes, algorithms, and more . Create more models through machine learning to make predictions more detailed and accurate, and automate data analysis through intelligent systems .
"Combine statistics, data analysis, machine learning" to use data to "understand and analyze actual phenomena". In other words, good data combined with a good trained model yields better predictions. The technology industry is rapidly evolving with the help of data science and making full use of predictive and prescriptive to forecast future trends.
4.0 Era KaaS (Knowledge as a Service) Knowledge as a Service
Rules Support Decisions
Data Analysis 4.0 → Rules Support Decision-Making
There are four main types of analysis: descriptive , elaborating on the past; diagnostic, using past data to study the present; predictive, predicting the future through insights based on past data; prescriptive, using models to guide optimal behavior.
Although data analysis 3.0 includes all the above types, it emphasizes the last one. While enterprises are pursuing data benefits, the machine model also exposes its shortcoming in analyzing business scenarios—the black box problem. Decision makers cannot be made aware of the process and reasons for analyzing and predicting problems.
Data scientists began to think about how to solve this key problem. After exploring and practicing various methods, they found that the innovative technology of logic rule fusion machine learning can break through the black box bottleneck.
It not only allows the machine model to maintain the ability of learning and prediction, but also has a certain degree of interpretability.
There is no doubt that artificial intelligence, machine learning, and deep learning have had a profound impact on the current era . Functions such as machine translation, smart reply, chatbots, meeting assistants, etc. will be widely used in the next few years. Data mining technology and machine learning algorithms have achieved a lot of results, and rule-supported decision-making will become a new stage of decision-making analysis .