Today, more and more important for business data analysis. From 2007 began in, Gartner began investigating the real use proportional share of personnel data analysis across the enterprise. In the past for a long period of time, the rate remained at 24% -25% or so, while in 2016 year 2017 , this proportion has gradually risen to 35% - the change indicated that currently there are about one-third of enterprises a person will make decisions with data and analysis.
Enhanced analysis is what?
In fact, experts or staff in the field of data analysis is to make data analysis more widely used for the public. If a company's employees can do data analysis, or everyone can make decisions based on data, then the business will be more successful, according to corporate decision-making will be more to follow.
However, most of the data analysis was based on an analysis of the enterprise known data, the data has produced some value, but also has done modeling. But there are a considerable number of unknown data has not been analyzed or modeled. For the analysis of these data, the traditional methods ineffective line. Therefore, Gartner introduced the concept of "enhanced analysis", and that it is the 2019 one in ten data to analyze trends.
The main application of enhanced scene analysis is to play real value of these unknown data and help users quickly find answers to known issues. The main starting point is to find the problem from unknown data to help enterprise users really dig out the value of the data yet to play.
Gartner research director at Sun Xin said, the enhanced analysis of "Data Preparation" can be analyzed in the data, "data mining models and modeling", "Share data discovery" gave three stages to bring the value of the user.
In the data preparation phase, general business can take advantage of the enhanced analysis behind the machine learning technology, the system automatically recommended for users of some of the data source, database, thus greatly reducing preparation time data.
In data mining models and modeling phase, the enhanced analysis can bring the recommended type algorithm is very easy to use features for the user. By simple single "drag pull", with can implement complex data analysis households, including prediction, clustering. Because of its low barriers to entry, business people can do very complex nature of data analysis by this method.
In the discovery phase of data sharing, enhanced insight and analysis can automatically bring true knowledge to the user. These automatic true knowledge and insight, automated data and found to have a higher reference resistance without prejudice. In addition, the operation will involve a lot of data analysis reports, and enhanced analysis can automatically tell the user which reports are important and which secondary, and the user can also use natural language to communicate with the system.
At present, the traditional large-scale data analysis vendor has its own flagship products, such as Cognos , SSAS / SSRS . Microsoft Power Bi and Tableau is now a modern business intelligence market leader, data analysis to visualize the form. Gartner believes that in the next two to five years, the market will have a lot of companies started to engage in enhanced analytics, allowing users to lower the threshold to do more advanced data analysis.
Sun Xinjiang stressed that this trend does not mean that high-level data scientists or statisticians companies will lose their jobs, and means that companies will be able to put these professionals saved time, to engage in more strategic things - let business staff more engaged in high-end analytics, and advanced data scientist or senior data analyst more involved in business decisions.
Enhanced analysis of what it can do?
Throughout the analysis, the enhanced analysis not only can achieve rapid preparation of data, automatic generation of insight, insight into what data can also tell the business or share useful and reduce their data analysis time. In addition, the enhanced analysis can help companies achieve high-level scenarios. For example, modeling takes a lot of manpower and time, but the enhanced automation of the analysis model can achieve the establishment of better management model.
Analysis can also be enhanced embedding mode. For example, business people can only do so in the traditional gateway to look at some of the basic work chart, but can now be enhanced analysis of traditional embedded platforms, the system then in the form of natural language to communicate, so the opportunity to do the analysis within the enterprise the proportion of employees rose to 50% .
In addition, traditional BI after visual graphics (business intelligence) system similar to the chart in the application of predictive models or clustering model module, business users can directly use these advanced analysis capabilities. More importantly, enhanced analysis brings the biggest advantage is "Pratt & Whitney Advanced analysis capabilities." In the past, advanced analytics just some of the expertise of professionals, but now all operational staff can operate, and with enhanced data analysis to reduce prejudice, inform the user of the need to focus on the real concern in an objective form.
Sun Xin said that enhanced analysis are applied in case banking, agriculture, insurance, retail and other business scenarios. For example, he said that in the banking sector, enhanced customer analysis which can detect the most noteworthy portrait. For example, the sale of traditional financial products often target older wealthy population; but through enhanced analysis, banks will find 25 Dao 35 years of age but is really adoption groups financial products.
Another example is in the insurance industry, the overseas sale of insurance policyholders based on specific attributes pricing . For example, the insured may have a sum of money each year to call an ambulance, but enhanced analysis will find that 12 children under the age of the insured amount but more needs to be given, because the ambulance on the needs of children accompanied by their parents, while the ambulance was outside by capitation fee.
These scenarios rely on general data analysis in fact can not be found. However, the enhanced form of analysis can be used machine learning to take into account all the conditions that may occur. These are the obvious benefits of enhanced analysis can bring.
Enhanced analysis of prospects
Sun Xin said that the enhanced analysis can be roughly understood as an enhanced version of the BI , but enhanced analysis BI provides a platform for deeper and stronger analytical capacity - Traditional BI doing mainly graphic chart, it may not be able to answer business questions , but the analysis can be enhanced.
However, Sun Xin do not think that type of analysis can be enhanced in the future instead of BI . "Because BI there is a very important business scenario is fixed reports, but this feature is an enhanced analysis of the current can not be achieved. And BI coverage is very wide, if the business people really need to be analyzed, then they need some flexibility enhanced analysis may be directly given to the results, during the lost flexibility, so many people will be analyzed in the form of programming - the highest flexibility, the enhanced analysis is not able to completely replace traditional BI role . "
But based on the obvious advantages of enhanced analysis brings many, Sun Xin believes that the next two to five years, the enhanced analysis will become the dominant market trend. If the platform does not have the enhanced analytical capabilities, enterprises will likely lose market competitiveness. Enhanced analysis can lower the threshold for allowing users to begin data analysis, and allows users to more sophisticated data analysis on a lower threshold.
In addition, the enhanced analysis as an analysis model based on machine learning, will become more powerful, the user can answer more questions. " In fact, from the manufacturer 's point of view, they hope that their products can be more enterprises to use, improve product sales. Enhanced positive analysis can lower the threshold to allow more users to do the analysis, so manufacturers are more willing to use enhanced analysis of the form of open market. "
