What Are the Different Types of Business Intelligence Technologies?

The concept of business intelligence was first proposed by the Gartner Group in 1996. The Gartner Group defines business intelligence as: Business intelligence describes a series of concepts and methods to assist business by applying fact-based support systems Decision making. Business intelligence technologies provide technologies and methods that enable companies to quickly analyze data, including collecting, managing, and analyzing data. Turn this data into useful information and distribute it across the enterprise.

Business Intelligence Technology

This entry lacks an overview map . Supplementing related content makes the entry more complete and can be upgraded quickly. Come on!
The concept of business intelligence was first proposed by the Gartner Group in 1996. The Gartner Group defines business intelligence as: Business intelligence describes a series of concepts and methods to assist business by applying fact-based support systems Decision making. Business intelligence technologies provide technologies and methods that enable companies to quickly analyze data, including collecting, managing, and analyzing data. Turn this data into useful information and distribute it across the enterprise.
Chinese name
Business Intelligence Technology
Nature
Support business decision making
Make up
Data warehouse data analysis, etc.
Purpose
Better business decisions
Business intelligence is the process of collecting, managing, and analyzing business information. The purpose is to enable decision makers at all levels of the enterprise to gain knowledge or insights that will enable them to make decisions that are more beneficial to the enterprise. Business intelligence generally consists of data warehousing, data analysis, data mining, online analysis, data backup and recovery.
At present, the definition of business intelligence in academia is not uniform. Business intelligence is generally understood as a tool that transforms existing data in the enterprise into knowledge and helps companies make informed business operations decisions. The data discussed here includes data from orders, inventory, transaction accounts, customers and suppliers from the industry and competitors of the enterprise and various data from other external environments in which the enterprise is located. Business intelligence decisions that can be assisted by business intelligence can be both operational and strategic and strategic. To transform data into knowledge, technologies such as data warehousing, online analytical processing (OLAP) tools, and data mining are needed. Therefore, from a technical perspective, business intelligence is not a new technology, it is just a comprehensive application of data warehouse, OLAP and data mining technologies.
It can be considered that business intelligence is the process of collecting, managing, and analyzing business information. The purpose is to enable decision makers at all levels of the enterprise to gain knowledge or insight. =, Prompting them to make decisions that are more beneficial to the business. Business intelligence generally consists of data warehouse, online analysis and processing, data mining, data backup and recovery. The realization of business intelligence involves software, hardware, consulting services and applications. Its basic architecture includes three parts: data warehouse, online analysis and processing, and data mining.
Therefore, it should be appropriate to consider business intelligence as a solution. The key to business intelligence is to extract useful data from many data from different enterprise operating systems and clean them to ensure the correctness of the data, and then extract, transform, and load, that is, The ETL process is merged into an enterprise-level data warehouse to obtain a global view of the enterprise data. Based on this, it is analyzed and processed using appropriate query and analysis tools, data mining tools, OLAP tools, etc. (at this time Information becomes knowledge to assist decision-making), and finally present the knowledge to the manager to provide support for the manager's decision-making process.
Well-known IT vendors providing business intelligence solutions include Microsoft, IBM, Oracle, Microstrategy, BusinessObjects, Cognos, SAS, etc. Domestic mainstream business intelligence vendors include SmartBI, FineBI, Qlikview, Yonghong Z-Suite, etc.
Business intelligence is defined as a collection of software tools
End-user query and reporting tools. Designed to support access to raw data for novice users, excluding finished product report generation tools for professionals.
OLAP tools. Provide multi-dimensional data management environment, its typical application is the modeling of business problems and business data analysis. OLAP is also called multidimensional analysis.
Data Mining (DataMining) software. Use techniques such as neural networks and rule induction to discover relationships between data and make data-based inferences.
Data Warehouse (DataWarehouse) and Data Mart (DataMart) products. Pre-configured software that includes data conversion, management, and access, and often includes some business models, such as financial analysis models.
The concept of online analytical processing (OLAP) was first proposed by EFCodd, the father of relational databases, in 1993. He also proposed 12 criteria for OLAP. The proposal of OLAP caused a lot of repercussions. As a type of product, OLAP is clearly distinguished from online transaction processing (OLTP).
Today's data processing can be roughly divided into two categories: On-Line Transaction Processing (OLTP) and On-Line Analytical Processing (OLAP). OLTP is the main application of traditional relational databases, mainly for basic, daily transaction processing, such as bank transactions. OLAP is the main application of the data warehouse system, supports complex analysis operations, focuses on decision support, and provides intuitive and understandable query results.
OLAP is a type of software technology that enables analysts, managers, or executives to quickly, consistently, and interactively access information from multiple perspectives, thereby gaining a deeper understanding of the data. The goal of OLAP is to meet decision support or to meet specific query and reporting needs in a multi-dimensional environment. Its core technology is the concept of "dimensionality".
"Dimension" is the angle from which people observe the objective world, and it is a high-level type division. "Dimensions" generally contain hierarchical relationships, which are sometimes quite complex. By defining multiple important attributes of an entity as multiple dimensions, users can compare data on different dimensions. Therefore, OLAP can also be said to be a collection of multidimensional data analysis tools.
The basic multi-dimensional analysis operations of OLAP are drill (rollup and drilldown), slice (slice) and dice (dice), as well as pivot, drillacross, drillthrough and so on. Drilling is to change the level of dimensions and transform the granularity of analysis. It includes rollup and drilldown. Rollup is to summarize low-level detailed data to high-level summary data in a certain dimension, or reduce the number of dimensions; while drilldown is the opposite, it goes from summary data to detailed data to observe or add new dimensions. Slicing and dicing are concerned with the distribution of metric data in the remaining dimensions after selecting values in a part of the dimensions. If there are only two dimensions, they are slices; if there are three, they are diced.
Rotation transforms the direction of a dimension, that is, rearranges the placement of dimensions in a table (such as row and column swapping).
There are multiple ways to implement OLAP, which can be divided into ROLAP, MOLAP, and HOLAP according to the different ways of storing data.
ROLAP stands for Relational OLAP. With the relational database as the core, the multi-dimensional data is represented and stored in a relational structure. ROLAP divides the multi-dimensional structure of a multi-dimensional database into two types of tables: one is the fact table, which stores data and dimension keywords; the other is the dimension table, which uses at least one table for each dimension to store the hierarchy of the dimension, Descriptive information for member categories and other dimensions. The dimension table and the fact table are linked by the primary and foreign keywords, forming a "star pattern". For dimensions with complex hierarchies, in order to avoid redundant data occupying too much storage space, multiple tables can be used to describe it. The extension of this star schema is called "snowflake mode".
MOLAP stands for MultidimensionalOLAP. Multidimensional data organization is the core, that is, MOLAP uses multidimensional arrays to store data. Multidimensional data will form a "Cube" structure in storage. In MOLAP, the "rotation", "cutting", and "slicing" of "Cubes" are the main techniques for generating multidimensional data reports.
HOLAP stands for Hybrid OLAP implementation (HybridOLAP). If the lower layers are relational, the upper layers are multidimensional matrix. This way has better flexibility.
There are other ways to implement OLAP, such as providing a dedicated SQLServer, and providing special support for SQL queries for some storage modes (such as star and snow flake).
OLAP tools are online data access and analysis for specific problems. It analyzes, queries, and reports data in a multidimensional way. Dimensions are the specific angles at which people observe data. For example, when a company considers the sales of a product, it usually observes the sales of the product in depth from different perspectives of time, region, and product. The time, region and product here are dimensions. The multi-dimensional array composed of different combinations of these dimensions and the measured indicators is the basis of OLAP analysis, which can be expressed formally as (dimension 1, dimension 2, ..., dimension n, metric), such as (region, time , Product, sales). Multi-dimensional analysis refers to taking various analysis actions such as slice, dice, drill-down and roll-up, and pivot on the data organized in a multi-dimensional form to analyze the data , So that users can observe the data in the database from multiple angles and sides, so as to understand the information contained in the data.
According to the different ways of organizing comprehensive data, there are currently two common OLAPs: MOLAP based on multidimensional databases and ROLAP based on relational databases. MOLAP organizes and stores data in a multidimensional manner, while ROLAP uses existing relational database technology to simulate multidimensional data. In data warehouse applications, OLAP applications are generally front-end tools for data warehouse applications. At the same time, OLAP tools can also be used in conjunction with data mining tools and statistical analysis tools to enhance decision analysis functions.
Summary At present, many vendors are active in the field of business intelligence (hereinafter referred to as BI). In fact, BI products and solutions that can meet user needs must be built on a stable and integrated platform. The platform needs to provide user management, security controls, connect data sources, and access, analyze, and share information. The standardization of the BI platform is also very important, because it is related to the compatibility problem with various application systems of the enterprise. If the compatibility problem cannot be solved, the BI system cannot play its due effect. Here we introduce the BI system by performing a functional anatomy of a laboratory BI system model (we call it the D system).
Decision support systems usually consist of two parts: data collection integration, data presentation and analysis. The former collects and integrates various data sources (ERP, OA, etc.) of the enterprise in multiple dimensions and angles for further analysis; the latter uses tailor-made data extraction, conversion and import tools to establish data warehouses or data market. On the basis of the data mart, various analysis models such as sales, purchase settlement, storage and transportation, finance, cost, budget, profit and loss, performance, and customers are established. These models are stored in a multidimensional database. Rotating perspective tables and other methods provide information demanders at all management levels of the enterprise to realize business intelligence. At the same time, third-party tools can be embedded to easily and quickly transfer decision support information to WAP users, PDAs, and mobile PCs to the hands of information users and final managers.
Business intelligence systems can assist in establishing an information center, such as generating various work reports and analysis reports. Used for the following analysis:
Sales analysis mainly analyzes various sales indicators, such as gross profit, gross profit margin, cross-ratio, sales-to-ratio, profitability, turnover rate, year-on-year, ring-to-month ratio, etc., while analysis and maintenance can be performed from management structure, category brand, date, time period, etc. From a perspective perspective, these analysis dimensions use multi-level drilling to obtain a fairly thorough analysis idea; at the same time, analysis data such as forecast information and alarm information are generated based on mass data; new perspective tables can also be generated based on various sales indicators.
Commodity analysis The main data of commodity analysis comes from sales data and commodity basic data, so as to generate analysis ideas based on the analysis structure. The main analysis data includes the product category structure, brand structure, price structure, gross profit structure, settlement method structure, and origin structure, etc., resulting in product breadth, product depth, product elimination rate, product introduction rate, product replacement rate, key products, best-selling Commodities, slow-selling products, seasonal products and other indicators. The analysis of these indicators by the D system guides the adjustment of the company's commodity structure, and strengthens the competitiveness and reasonable allocation of the commodities it operates.
Personnel analysis The company's personnel indicators are analyzed through the D system, especially sales personnel indicators (mainly sales indicators, supplemented by gross profit indicators) and purchasing personnel indicators (sales, gross profit, supplier replacement, number of purchased and sold products, and consignment products). Data, capital occupation, capital turnover, etc.) in order to assess employee performance, improve employee motivation, and provide a scientific basis for the rational use of human resources. The main topics of analysis are: staff composition, sales per capita of sales staff, personal sales performance for sales, per capita sales of various management structures, gross profit contribution, how much purchasers control the purchase of goods, the proportion of purchase and sales consignment, introduction How much of your product sales and so on.
The implementation of business intelligence systems is a complex systems project. The entire project involves knowledge in many categories such as enterprise management, operation management, information systems, data warehouses, data mining, statistical analysis, and so on. Therefore, in addition to choosing the right business intelligence software tools, users must follow the correct implementation method to ensure the success of the project. The implementation steps of business intelligence projects can be divided into:
(1) Requirements analysis: Requirements analysis is the first step in the implementation of business intelligence. Before other activities, the expectations and requirements of business intelligence must be clearly defined, including the topics to be analyzed, and the perspectives (dimensions) that each topic may view ; Need to discover the laws of those aspects of the enterprise. The needs of users must be clear.
(2) Data warehouse modeling: Through analysis of enterprise needs, establish logical and physical models of the enterprise data warehouse, and plan the system's application architecture to organize and classify various types of enterprise data according to the analysis theme.
(3) Data extraction: After the data warehouse is established, the data must be extracted from the business system into the data warehouse. During the extraction process, the data must also be converted and cleaned to meet the needs of analysis.
(4) Establishment of business intelligence analysis reports: Business intelligence analysis reports need to be developed by professionals in accordance with the format formulated by users, and users can also develop them by themselves (the development method is simple and fast).
(5) User training and data simulation test: For the development-use of a separate business intelligence system, the use of the end user is quite simple, and only a click operation can be performed to analyze specific business problems.
(6) System improvement and perfection: The implementation of any system must be continuously improved. The business intelligence system is even more so. After users use it for a period of time, more and more specific requirements may be put forward. At this time, the system needs to be restructured or improved according to the above steps.
In a changing market environment, companies must strive to become the dominant player, not a follower. The biggest benefit of business intelligence is that it can get the most accurate and timely information to help companies gain a competitive advantage. Enterprise decision makers can analyze customer consumption trends, cultivate loyal customers, strengthen connections with suppliers, reduce financial expenditures, tap new business opportunities, analyze future development trends, develop business strategies, adjust product structures, and distribution channels. , Workflow and service methods.
Business intelligence involves complex processes and many related technologies. Its architecture can adapt to increasing workloads and user needs. Business intelligence can "fundamentally help you transform your company's operational data into high-value, accessible data." The right information or knowledge, and the right information to the right people in the right way at the right time. "
In practice, it can be applied in many forms. We know that business intelligence can help us find the information we need and share and analyze it. The main applications of business intelligence are:
(1) Integration of internal and external information
This information can be data stored in databases, data marts, data warehouses, or unstructured arbitrary files (HTML, text, electronic) generated in enterprise application systems such as SCM, CRM, ERP, BPR, SFA, etc. Forms, DOC, etc.).
(2) Simple reports and queries
At this level, business intelligence is simply the rough processing of information. What was the sales volume of the Y product last month? How many customers do we have? The question the user asked the system was "tell me what happened."
(3) Online analysis and processing (OLAP)
</ strong> Business intelligence tools enable users to analyze information, create value-added information and better integrate information. In which region has our product achieved the greatest success? Which part of the user purchased what kind of product? Based on a simple report, we can further ask many questions. Ask not only what happened, but also why.
(4) Manager Information System (EIS)
At present, the basic means for leaders to obtain business information is reports. For various reasons, the reports prepared solely by the information department cannot fully meet the leaders' needs for information, and consulting a large number of corporate reports will take up a lot of time for senior leaders. The high-level leadership information system established on the basis of business intelligence analysis software can help leaders at all levels in the industry to obtain information quickly and easily. At the same time, the active query mechanism of related tools can be used to query key data related to industry business in the information data. If an abnormal operation of the industry is found through the query, the system will automatically report to leaders through various means to remind relevant leaders Pay attention to the problems that arise.
(5) Data drilling
Through statistical methods, the future scene can be shown in detail. For example, by using business intelligence tools, we can predict which customers are most likely to buy our new products. Marketing strategies can thus focus on a limited number of customers. As a result, the company's marketing strategy is more effective and costs can be reduced. In this case, our question is: tell me what will happen in the future.
(6) Regional information support
</ strong> The business intelligence platform is not only confined within the enterprise, but can be extended to a relatively large area, allowing more users to share information. Information may be stored in different databases in the organization, or outside the organization, and may include sound and images. Business intelligence can provide fast, timely, and accurate information to users of enterprises and organizations in any region and at the right time, thereby greatly improving their judgment and decision-making level.
(7) Monitoring basic performance indicators of enterprises
The basic performance indicators of an enterprise are calculated by "extraction" from enterprise application systems, which include sales, marketing, customer service, finance, human resources, manufacturing, and supply chain. The business intelligence system's performance information architecture can set corporate goals based on basic indicators, compare real-time conditions with planned values in real time, and suggest possible contingency actions. Through the performance information architecture, the system can calculate business intelligence information, provide business executives with instant information, and business executives can further analyze and query overall or detailed information.
(8) Mining business rules
</ strong> The data mining method is used to extract the implicit, previously unknown, and potential application information from the database. Data mining differs from traditional analysis tools in that data mining uses a discovery-based approach, using pattern matching and other algorithms to determine the important relationships between data.
Business intelligence technology is a complete information supply chain that connects information producers and users. It can enable the staff in the organization to use appropriate tools to obtain enterprise information according to their needs, thereby enhancing the core competitiveness of the enterprise.
This new supply chain relationship allows information technology personnel to make full use of their advantages in database management, security assurance, infrastructure maintenance, report writing and distribution, etc., while enabling information users to rely on their own business knowledge and focus In answering business questions.
Although the steel trade has a large amount of basic transaction data, today's systems can only perform simple statistical analysis, and cannot provide targeted reports and analysis for decision-making. As a result, many managers spend 80% of their time on data analysis, and only 20% of their time is used for decision-making. At the same time, a large professional analysis team must be equipped for a large number of analysis work.
Business intelligence is proposed on the basis of information management systems such as ERP. It is an intelligent management tool based on information technology. It performs various analyses and reports on enterprise data generated by systems such as ERP, CRM, and SCM. Help managers understand the status of enterprises and markets and make the right decisions.
(1) BI caters to the needs of enterprises. BI can calculate a large amount of information, and then find out market trends, business problems, and new market opportunities for enterprises.
(2) The decision support effect provided by BI to enterprises is obvious. Because BI is based on the analysis of the primary data of the enterprise, the decision-making effect on the enterprise is significant.
(3) BI allows all employees of the company to analyze the data themselves to find problems. This is a function that some original decision support systems did not have. Generally speaking, the BI system will categorize a large amount of information to bring related information together, such as finance, inventory, quality and other topics. When you need specific information, you can use BI software to search for the information you need.
(4) Data collection and data mining have provided a good foundation for BI, and corporate demand for BI has increased. Decision makers hope to be able to use existing data to guide corporate decisions and improve their competitive advantage. Such data will bring significant added value if used for decision support. If external data such as industry analysis reports, independent market surveys, evaluation results, and consultant evaluations are added, the benefits of the above processing process can be further enhanced. Therefore, business intelligence has gradually become a hot spot in recent years. It can be said that it is inevitable that the development of enterprise informatization has reached a certain level. BI is based on data warehouse and uses OLAP tools and data mining technologies to provide support for corporate decision-making.
For the application of BI, almost every enterprise needs it. The following discusses the main application of BI in the field of steel trade based on the author's actual work experience.
(1) Sales analysis
</ strong> A daily summary of sales by region, department, salesperson, and product provides managers with comparison and trend analysis to help identify issues and opportunities. BI applications can analyze and evaluate the sales of past products to determine the factors of product success or failure. With DSS, company-wide data can be used to infer the profits and revenues implied by a decision.
(2) Market analysis
The market analysis and decision-making module includes: (1) analysis of market distribution; (2) analysis of market product competition; (3) analysis of the impact of price changes on demand; (4) analysis of opening new markets. To this end, the decision support system should provide a market potential model in order to support managers to consider improving product competitiveness, occupying unrealized markets, and opening up undeveloped markets. These models are used as support tools for sales decisions. They can be used to correctly select the target market and key markets of the company's products, formulate policies, preparations and strategies for opening up, occupying and expanding the market, and correctly formulate product price policies and promotion strategies. Economic benefits of production and operation activities.
(3) Customer analysis
</ strong> Decision support system applications can use statistical tools to analyze the transaction data collected daily to determine the consumption patterns of various types of customers, and then take corresponding marketing measures to achieve maximum profits. For key customers, we need to provide better services and more favorable pricing strategies; for potential customers, we must promote them; for customers who are easy to lose, we must analyze the reasons to recover.
(4) Market research
Market research includes: using predictive model analysis to derive the growth model of each product in order to make appropriate decisions to terminate or expand a certain product; corporate brand and image research in order to increase the visibility and reputation of the company and brand; analysis of customer satisfaction Degree; market size and potential size research.
(5) Financial analysis
Compare actual costs and expenses by year, month, day, or other custom cycles; review past cash flow trends and predict future cash requirements; budget planning and cost allocation for complex projects; integrate finance across branches Data to form correct and consistent financial statements.
(6) Enterprise analysis
</ strong> Business analysis includes profitability, current ratio, market share, human resources analysis, but also advertising analysis, product pricing, weekly (daily) sales results, customer sales potential, market trends, foreign trade and exchange rates, Industry trends, labor cost trends, etc.
(7) Product cost analysis
</ strong> Generate product cost analysis data sources based on the analysis dimension data and analysis entity data of the product purchase price, sales price, product shipping cost, storage fee, other management costs, capital cost expense, sales volume, etc. Statistics on the cost of various products. Use these basic analysis data to flexibly generate various types of procurement analysis reports, and use drilling, turning and rotating analysis methods to analyze other interesting results of finding finished products.
(8) Price analysis
Use the model provided by the system or the user-defined function to analyze the price fluctuation ratio, classify products with a large increase or decrease in price within a certain period of time, and analyze the causes of price fluctuations in order to timely and effectively formulate or adjust marketing, procurement, Inventory policy. Price warning analysis, observing the highest and lowest prices of a product exceeding the height of the warning line, and evaluating the proportional relationship between the price and the predetermined price, and timely grasp the abnormal situation of the market price of the product.
Business intelligence helps the company's management to make fast and accurate decisions, quickly finds problems in the enterprise, and prompts management to resolve them. However, business intelligence software systems cannot make decisions on behalf of managers and cannot automatically deal with problems encountered in the operation of enterprises. Therefore, business intelligence systems cannot bring direct economic benefits to enterprises, but it must be seen that business intelligence brings businesses It is a scientifically armed management thinking that brings the entire enterprise fast and accurate decision-making, the timeliness of problem discovery, and the discovery of potential knowledge and laws that those opponents have not discovered, and this information is the enterprise The basis for generating economic benefits. Failure to quickly and accurately specify a decision-making policy is equivalent to sending the market to an opponent. Failure to discover potential information in a business in time is a waste of your resources. For example, through the analysis of sales data, you can find the relationship between the characteristics of various types of customers and the products that they like to buy, so that more targeted and accurate promotions can be carried out or more personalized services can be provided to customers. Bring direct economic benefits to the enterprise.
Manufacturing is an important market for business intelligence
According to a report by ManufacturingInsights (a subsidiary of IDC), the manufacturing IT market in the Asia-Pacific region (excluding Japan) was US $ 13.7 billion in 2004. It is expected that the market will grow steadily at a compound annual growth rate of 11.4%. By 2008, the market size It will reach $ 21 billion. At the end of 2004, IT spending in the Asia-Pacific region (excluding Japan) totaled $ 13.7 billion, of which discrete manufacturing accounted for 78.6% and process manufacturing accounted for 22.4%. As market globalization and liberalization bring more intense competition and complexity, many manufacturers in the Asia-Pacific region (excluding Japan) continue to invest in IT to improve operational efficiency and better control growing business costs. As more and more manufacturers set up production bases in China, reducing costs and occupying a huge domestic market, these manufacturers need to invest in major IT infrastructure, applications and services to enable their operations to develop healthily and steadily And gain a lead. This will continue to promote manufacturing IT investment by Chinese and overseas manufacturers. While investing heavily in infrastructure, many manufacturers in emerging large markets like China and India will continue to invest in enterprise resource management (ERM) and business intelligence (BI) solutions to better internal collaboration And decision-making provide a basic platform.
IDC report shows that the business intelligence (BI) tool software market size in Asia Pacific (excluding Japan) was US $ 233.2 million in 2004. It is expected that the market will grow rapidly at a compound annual growth rate of 12.3%. By 2009, the market size will reach At $ 417.3 million, the growth is expected to come mainly from the growing economies of China and India. The healthier economic environment and increasing application system deployment in these two countries have laid the foundation for the adoption of BI tools in the next five years. Relevant experts pointed out that with the popularity of the Internet, it has become inevitable to develop business intelligence based on decision support systems. With the application of various Internet-based information systems in enterprises, enterprises will collect more and more information about customers, products and sales, which can help enterprises better predict and grasp the future. Therefore, the development of e-commerce has also promoted the further application of business intelligence.
From the perspective of industry development, business intelligence is a business-driven decision support system, and its development is based on relatively complete enterprise information systems and stable business systems. The future application of business intelligence is closely related to the basic status of informatization in the industry, mainly manufacturing enterprises, followed by distribution companies. These two fields will be new markets for business intelligence that cannot be ignored. With the improvement of information technology, business intelligence products will be further integrated with management software such as ERP and CRM. At present, many ERP vendors have embedded business intelligence into corresponding ERP systems. For example, SAP's ERP has nested BO companies. AD's business intelligence products, AD also has a similar cooperation.
Of course, business intelligence, like ERP, has certain risks in its implementation. Enterprises must first recognize their own needs and fully understand them when selecting partners. Each mainstream manufacturer has its own advantages, such as SAS data mining, Hyperion's budget and report consolidation, and BO's data analysis and reporting. The development trend of business intelligence products is bound to be integrated applications based on integrated platforms. How to truly understand your own needs and choose the products of manufacturers with advantages will be the key to the success of business intelligence implementation.
Compared with DSS and EIS systems, business intelligence has better prospects for development. In recent years, the business intelligence market has continued to grow. IDC predicts that by 2005, the BI market will reach $ 11.8 billion, with an average annual growth rate of 27% (InformationAccessToolsMarketForecastandAnalysis: 2001-2005, IDC # 24779, June2001). With the introduction of enterprise CRM, ERP, SCM and other application systems, enterprises do not stay in the transaction process and focus on the effective use of enterprise data to support accurate and faster decision-making. The demand for business The need for intelligence will be huge.
The development trend of business intelligence can be summarized as follows:
Functionally configurable, flexible and changeable
The scope of a BI system extends from serving specific users in a department to serving all users across the enterprise. At the same time, due to differences in corporate users' powers and requirements, BI systems provide a wide range of targeted functions. From simple data acquisition to the use of WEB and LAN and WAN for rich interaction, analysis and use of decision information and knowledge.
The solution is more open, extensible, and customizable, providing a customized interface while ensuring the core technology
In response to the unique needs of different enterprises, the BI system provides core technology while personalizing the system, that is, adding its own code and solution based on the original solution to enhance the customized interface and expansion characteristics; Provide customized tools based on business intelligence platform to make the system more flexible and useable.
From standalone business intelligence to embedded business intelligence
This is a major trend of current business intelligence applications, that is, embedding business intelligence components in the existing application systems of enterprises, such as financial, human, and sales systems, so that transaction processing systems in the general sense have the characteristics of business intelligence. It is not a simple matter to consider a component of a BI system instead of the entire BI system, such as applying OLAP technology to an application system, a relatively complete business intelligence development process, such as enterprise problem analysis, solution design, and prototype system development , System application and other processes are indispensable.
Transition from traditional functions to enhanced functions
Enhanced business intelligence functions are relative to earlier business intelligence functions that used SQL tools to implement queries. In addition to the traditional BI system functions, most of the currently used BI systems have implemented the functions of the data analysis layer in Figure 2. Data mining and enterprise modeling are applications that BI systems should strengthen to better improve system performance.
After several years of accumulation, most of the large and medium-sized enterprises and institutions have established relatively complete basic information systems such as CRM, ERP, and OA. The unified characteristics of these systems are: through the operation of business personnel or users, the database is finally added, modified, and deleted. The above systems can be collectively referred to as OLTP (OnlineTransactionProcess, online transaction processing), which means that after the system has been running for a period of time, it will certainly help enterprises and institutions to collect a large amount of historical data. However, the large amount of data that is dispersed and independent in the database is just a few heavenly books for business personnel. What business people need is information, abstract information that they can understand, understand, and benefit from. At this time, how to transform data into information so that business personnel (including managers) can fully grasp and use this information and assist decision-making is the main problem that business intelligence solves. How to turn the data in the database into the information needed by business people? Most of the answers are reporting systems. Simply put, the reporting system can already be called BI, it is a low-end implementation of BI.
Now most of the foreign companies have entered the mid-end BI, which is called data analysis. Some companies have started to enter high-end BI, called data mining. Most of the enterprises in China are still in the reporting stage.
Data reports cannot be replaced
The traditional reporting system is technically quite mature, and everyone is familiar with Excel, Crystal Reports, ReportingService and so on. However, with the increase of data and the increase of demand, the traditional reporting system faces more and more challenges.
1. Too much data and too little information
Dense tables are piled with a large amount of data. How many business people look at each data carefully? What information and trends do these data represent? The higher the level of leadership, the more concise information is needed. If I were the chairman, I would probably only need one sentence: Is our situation good, medium or poor?
2. Difficult to analyze and understand various combinations interactively
The customized report is too rigid. For example, we can list the sales of different regions and different products in one table, and the sales of customers in different regions and different ages in another table. However, these two tables cannot answer questions such as "the situation of young and middle-aged customers in North China purchasing digital camera-type products". Business problems often require interactive analysis from multiple perspectives.
3. Difficult to dig out potential rules
The reporting system often lists data information on the surface, but what rules are potentially contained in the depth of massive data? What customers are the most valuable to us and how well are the products connected to each other? The greater the value, the harder it is to dig out.
4. Difficult to trace history, data form silos
There are many business systems, and data exists in different places. Too old data (such as data from one year ago) is often backed up by business systems, making macro analysis and long-term historical analysis difficult.
Therefore, with the development of the times, the traditional reporting system can no longer meet the growing business needs, and enterprises are looking forward to new technologies. The era of data analysis and data mining is coming. It is worth noting that the purpose of data analysis and data mining systems is to bring us more decision support value, not to replace data reports. The report system still has its irreplaceable advantages, and will coexist with the data analysis and mining system for a long time. Data analysis in more than eight dimensions
If OLTP focuses on daily transaction operations such as adding, modifying, and deleting databases, OLAP (Online Analysis Process) focuses on analyzing macro issues and comprehensively analyzing data to obtain valuable information.
In order to achieve the purpose of OLAP, the traditional relational database is not enough. A new technology called multidimensional database is needed.
The concept of multidimensional databases is not complicated. As an example, we want to describe when Cola sold 100,000 yuan in the northern region in April 2003, several angles were involved: time, product, region. These are called dimensions. As for sales, they are called measures. Of course, there are costs and profits.
In addition to time, product, and region, we can also have many dimensions, such as the gender of the customer, occupation, sales department, promotion method, and so on. In fact, the multidimensional database in use may be an 8- or 15-dimensional cube.
Although the 15-dimensional cube is complex in structure, it is conceptually very simple.
The overall architecture of the data analysis system is divided into four parts: source system, data warehouse, multidimensional database, and client.
Source system: Including all existing OLTP systems, building a BI system does not require changes to the existing system.
Data warehouse: Data is concentrated. Through data extraction, the data is continuously extracted from the source system. It may be once a day or once every 3 hours. Of course, it is automatic. Data warehouses are still built on relational databases and often conform to a model called a "star structure."
Multi-dimensional database: The data of the data warehouse is multi-dimensionally modeled to form a cube structure. Each cube describes a business topic, such as sales, inventory, or finance.
Client: Good client software can colorfully display the information in the multi-dimensional cube to the user.
Data analysis case:
In the actual case, we use Oracle9i to build a data warehouse, Microsoft AnalysisService2000 to build a multidimensional database, and ProClarity6.0 as the client analysis software.
The decomposition tree looks like an organization chart. Decomposition trees are effective in answering the following questions:
Which product has the highest sales within the specified product group?
What is the distribution of sales among various products within a specific product category?
Which salesperson completed the highest percentage of sales?
Data mining sees through your needs
Broadly speaking, any process of mining information from a database is called data mining. From this point of view, data mining is BI. However, technically speaking, data mining (DataMining) specifically refers to: the source data is cleaned and transformed into a data set suitable for mining. Data mining completes the extraction of knowledge on this fixed-form data set, and finally uses appropriate knowledge models for further analysis and decision-making. From this narrow perspective, we can define: Data mining is the process of refining knowledge from a specific form of data set. Data mining often targets specific data and specific problems. One or more mining algorithms are selected to find the rules hidden under the data. These rules are often used to predict and support decision making.
Related sales case:
American supermarkets have such a system: When you purchase a cart of goods and check out, after the saleswoman scans your product, some information will be displayed on the computer, and then the salesperson will kindly ask you: We have a disposable paper cup Promotions, located on F6 shelves, do you want to buy?
This sentence is by no means a general promotion. Because the computer system is long gone, if you have napkins, large bottles of cola and salad in your shopping cart, then 86% of the time you will buy disposable paper cups. As a result, you said, ah, thank you, I just couldn't find the paper cup just now.
This is not a magical scientific fortune-telling, but a system implemented using association rule algorithms in data mining.
Every day, new sales data will enter the mining model and be processed by the mining model along with the historical data of the past N days to get the current most valuable association rules. The same algorithm analyzes the sales performance of online bookstores, and the computer can find the correlation between products and the strength of the correlation.
Data reporting, data analysis, and data mining are the three levels of BI. We believe that the trend in the next few years is that more and more enterprises will enter the field of data analysis and data mining based on data reports. The decision support function brought by business intelligence will bring us more and more obvious benefits.
At present, the major business intelligence software vendors in the domestic market are: Kingdee, UFIDA, IBM, Power-BI, ORACLE, SAP, SAS, Sybase, Analyzer, Microsoft, Frenet, and Qin

IN OTHER LANGUAGES

Was this article helpful? Thanks for the feedback Thanks for the feedback

How can we help? How can we help?