Blue text Reading punch card

Book title: "Intelligent Business"

Chapter |:04 Data Intelligence

Book summary tips:

To integrate data intelligence into a specific business, you have to do three things well: data, algorithms, and products. (The trinity of data, algorithms, and products provides a service.)

Dataization:For business today, intelligence means that business decisions will increasingly rely on machine learning, on artificial intelligence. The biggest business value in the next 10 years is how to create an intelligent business that brings a leap in user experience.

The dataization of specific business scenarios, algorithms and their iterative optimization that are faithful to business logic, and products that seamlessly integrate data intelligence with business scenarios.

"Datafication" is essentially the process of transforming a phenomenon into a quantifiable form. It stems from the fact that the human desire to measure, record, and analyze the world is the foundation of civilization's progress. Only when we have a sufficiently large amount of multi-dimensional "big data" can we truly understand our surroundings, the origin of things, and ourselves objectively, truthfully, and profoundly.

Data initialization is costly and difficult. Effective data initialization is a critical first step in creating value from Big Data. Whether or not the costly data initialization work can create huge customer value becomes an important test of whether or not a huge number of startups can survive today.

Algorithmization: the "engine" rather than the "tool" of smart business. In a business context, an algorithm is a set of computational instructions that reflects the logic of the product and the market mechanism. An algorithm is a set of instructions that run according to a set program to obtain a desired result. After completing the dataization of the business scenario, the algorithm is the idea of refining the value of the data, and the value of the data is the business value.

Algorithms are the core of machine learning. Algorithmic models explore the data in real-time online, full-text records, without prediction or direction, to discover those widely latent but unnoticeable relationship structures, to find the hidden connections among thousands of possible factors, and to continuously optimize.

The industrial revolution automated manual labor, the information revolution automated mental labor, and machine learning automated the process of automation itself. Algorithms are at the heart of intelligence. Based on data and algorithms, machine learning is accomplished and artificial intelligence is realized.

Productization: Data closure is achieved through product interaction, and the product experience relies on data intelligence, where data and product become one. The success of smart business is often a result of defining a new way of user experience in response to a user problem, and at the same time starting the engine of data intelligence to continuously improve the user experience. Data, algorithms and productization are the "trinity" of intelligent commerce completed in a closed loop of feedback.

"Big data" is characterized by large volume, variety, speed, and high quality.

The concept of "live data": to connect data seamlessly with real life, data needs to be recorded online in real time, not actively collected; it needs to be constantly updated and available at any time to generate insights; and it needs to be used flexibly in real business scenarios to drive the next decision.

The characteristics of "live data" are:1. full record rather than sample survey. 2. data first, then insights. 3. data is decision-making.

The biggest premise of "live data" must be a significant reduction in the cost of data recording. If this premise can not be met, it can not be realized. In the era of "talking data", we value correlation rather than causality. The engine of data intelligence machine to be able to make decisions directly, rather than the traditional use of data analysis to support human decision-making.

Companies are big and small. From the perspective of "live data" to think, you will understand that the size of the data volume is only a relative concept, if you let the data become a natural part of your business, so that the machine to become a part of your decision-making, your business behavior will go into the fast lane of intelligence.

The so-called enterprise intelligence, simply put, is a "decision-making machine that can learn", not only to make decisions, even the efficiency and effectiveness of decision-making can also be optimized through the closed loop of learning to improve, which of course requires the support of data and algorithms.