Some questions about website data Analysis 2

The intermediary transaction SEO diagnoses Taobao guest stationmaster buys the Cloud host technology Hall

Previous--Web site data analysis of some of the problems in 1 mainly listed in the Web site data analysis industry and data analyst this career-related issues, this is the second, the main want to list some questions about BI.

Bi (Business FDI, Business Intelligence), first look at the definition of bi on Wikipedia:

Business FDI (BI) is tabbed as the ability for a from to take all it capabilities and convert them int o knowledge.

BI provides a wealth of valuable information to guide companies in finding new opportunities for development, and when companies recognize potential opportunities and successfully implement strategic decisions, BI can help companies build competitive advantages in the marketplace and sustain enterprise development. Bi is often associated with decision support systems (Decision Support system, DSS), in fact, the main goal of BI is to achieve decision support for the enterprise.

Here are a few questions about BI:

What is the relationship between Q1, Bi, and the Data Warehouse (DW)? (know)

The first thing to be clear is that BI focuses on the application of the data, making the data valuable, and all the underlying data is derived from the Data warehouse.

BI is defined in two directions: The broad bi is a data warehouse, the broad bi includes data acquisition, processing, storage, after analysis, mining, display into the whole process of valuable information, composed a complete system, Of course, in this system the Data warehouse takes the responsibility of processing and storing from the data acquisition, is the basic component; The narrow bi only includes the data application of the upper layer, including the data display, analysis, mining, etc., so it does not include the Data warehouse.

Because the definition of BI is more focused on data applications, and as the volume of data is not enlarged, the data warehouse is more isolated as a separate technology, so the current BI and data warehouse definitions are more likely to be separated, the entire system is called "Dw/bi" solution.

Q2, BI system is mainly to help enterprises solve what kind of problem? (know)

The objective of BI is to optimize the decision support of the enterprise, realize the transformation from data to valuable information, and assist the enterprise business strategy and decision making. So the ultimate goal of BI is to get the business insight.

Bi is the first to realize the transparency of enterprise data, the original data report is in order to quantify the data from the point of view of the operation of the enterprise, with the support of data, many decision-making will have a reference basis. With the development of business and information technology, BI is no longer just in the field of reporting, and data is being used for business analysis in addition to presentation, and the basis of business analysis is statistics, forecasting and optimization, which play a more critical role in enterprise's operational decision-making. But as the information expands and the volume of data increases, BI continues to face challenges, and we need to spend more to process and store data, and more energy to analyze and apply data. I have written about the three contradictions in BI's application. This article, because for a while, many places of view may have changed, but these 3 contradictions believe still exist.

Therefore, ultimately, it is necessary to grasp the output of BI is valuable information, regardless of the middle of the processing method is a query, report, or analysis, mining, ultimately to come to a valuable conclusion.

What are the main applications or components of Q3, current bi? (know)

Here briefly summed up, there may be omissions, I hope you can add in the comments. This only includes some functions based on the data application level in narrow bi, and the application of data processing in Data Warehouse is not listed here.

The first is the reports, the charts, and the dashboard, where the current reports and charts are richer, and the key difference from the traditional reports is interactivity. The present report basically all provides the simple data to filter, the sorting and so on function, the dashboard appearance has realized the on-demand consolidated report and the chart function.

Moreover, Olap,olap was once used as the core function of BI, and had to admit that OLAP is the most effective means of analyzing data, especially based on multiple dimensions of analysis, which are one or two of the chart can not do. OLAP is generally based on the multidimensional models that have been designed and the data marts that store multidimensional models, data marts and OLAP have many associations with the business level, which distinguishes the data mart from the underlying Data Warehouse (Mart).

Then the data query and analysis, sometimes based on the established model of OLAP can not meet the needs of analysis, so there is the need for data query, the general direct query data Warehouse details data; Hoc query in BI is a flexible query for a given multidimensional model, which allows you to combine dimensions and metrics freely.

Finally, the release of the report and data early warning, this is a bi platform for the push function, generally can be through the form of e-mail subscriptions to the combination of reports to the relevant personnel, and through the setting of early warning, you can monitor the trend of data changes, grasp the data may occur anomalies.

In addition, BI has many novel functions, such as map data based on GIS, dynamic chart based on flash and integration of data mining function.

What is the practical value of the multidimensional data model and OLAP in Q4, bi? (know)

Before the introduction of multidimensional data model and OLAP, you can refer to the Data Warehouse multidimensional data model and data cube and OLAP two articles in the content.

In fact, multidimensional data model and OLAP is the most important to solve the problem of how to effectively observe the data, the traditional relational model is difficult to directly analyze the data, and multidimensional model for the data Viewer provides a clear perspective, as usual we look at things from a variety of angles, The Multidimensional Model dimension design provides a good choice of these angles. And OLAP's several operation forms is manifests the "analysis" the word itself the meaning, from the overall to the detail, unifies several dimensions the cross analysis, lets us have the entire dataset to carry on the panoramic observation ability.

OLAP's most critical technology in addition to multidimensional model design and is expected to calculate (precomputation), or called pre-aggregation, is expected to solve the problem of rapid data acquisition, based on a certain rule or algorithm for the data set for the pre-calculation, OLAP operating performance may be effectively improved, This makes it possible to quickly and flexibly analyze and manipulate large amounts of data.

Q5, the current market mainstream BI products are mainly? (know)

The main business bi products in the market include IBM's Cognos, in addition IBM has its own DB2 can build Data Warehouse, in 2010 after the acquisition of SPSS, let it in the field of data analysis and data mining is also more competitive, SAP Business Objects (BO), In addition, SAP has BW (Business information Warehouse), as a traditional ERP solution provider has a unique advantage in data integration, Oracle BI (Enterprise-class called Biee,oracle Business FDI Enterprise Edition), Oracle has unique advantages in building data warehouses with its powerful relational database. These 3 big business bi are integrated BI, plus the SSIS, SSAS, and SSRS provided by Microsoft with SQL Server databases are also integrated BI solutions. There are also independent bi companies, such as SAS, with traditional advantages in data mining, Micro strategy BI solutions, open source strong bi Systems Pentaho (there are many open source BI systems in the previous years, but because BI has a technical threshold and cost, So at present many open source Bi will include open source version and commercial version, Pentaho is no exception, domestic also have Ufida BQ software is also belong to BI products.

Summed up is the current BI products mainly commercial products, and the whole set of BI products are generally heavyweight, in the purchase, deployment and use of a certain cost input.

If you have an opinion on BI, you are welcome to comment below or to answer the appropriate questions.