IBM-created large data analysis Empire

When you think of big data Analytics, you might think of startups. Please think again. IBM has long been involved in this area and may do better than any other company.

IBM's Deepak L.K.

IBM is making great success in large data areas. Its IBM Large Data website (http://www-01.ibm.com/software/data/bigdata/) is a large resource, its television advertising embodies the authority, its products in this area has established a strong position. But it seems to me that after a thorough conversation with IBM's VP of Business Analytics Products and solutions, Deepak L.K., I have a comprehensive understanding of IBM's Big Data strategy.

Whose new trick is best?

IBM is a company with a history of 101 years, headquartered on the east coast of the United States. It produced typewriters and is still making large machines. Its products can work in tandem with open source technology, but most products are not free at all. And many of those products come from a series of companies that the company has acquired in the course of its development. In addition, IBM is a service company with a large number of consultants all over the world.

In contrast, most of the emerging companies in the big data field have a streamlined team that was established just a few years ago, often headquartered on the west coast of the United States, providing core technology that is often open source software running on inexpensive, popular hardware and gradually developing intellectual property inside. However, IBM is still a big player in the big data and analytics field. How can it be that IBM has so many important statistics that seem to be out of tune with the field of contention? A conversation with L.K. reminded me of several of these reasons, and he pointed out several other reasons I didn't know. He helped me understand the relationship between each other and get a clearer picture of the ins and outs.

Real acquisition History

For example, I know that IBM has acquired a number of companies in the field of Business analytics over the past ten years. One of them is particularly easy to remember: IBM acquired a heavyweight company cognos--in the business intelligence sector in 2007, and earlier in the year it acquired Applix. The deal gives IBM an End-to-end business Intelligence suite that includes traditional and in-memory online analytical processing (OLAP), reports, dashboards, and data visualization. I know that.

But I'm not sure about any of the other acquisitions by IBM, but the combined results of all the acquisitions are what brought the value of integration analysis. Prior to the acquisition of Cognos, IBM also acquired the Ascential software company in 2005, thus bringing the extraction, transformation and loading (ETL) products DataStage and other assets to IBM. Prior to the takeover of Cognos, IBM acquired the statistics and analysis Heavyweight company SPSS in 2009, and in 2010 acquired a large-scale parallel processing (MPP) Data Warehouse professional manufacturer Netezza.

So in addition to the complete business Intelligence suite, IBM also has the tools necessary to perform the high performance data warehousing technology needed to enter data for those business intelligence systems and to implement predictive analysis of output results. Speaking of analysis, this is an area closely related to risk management. With this in mind, IBM's acquisition of Openpages and Algorithmics in 2010 and 2011 was an extremely powerful boost to its big data strategy.

These are just a few of the products IBM has in the field of Business Analytics. In fact it has a lot of products, as L.K. explained to me, IBM has a complete "Featured solution" program (http://ibmdatamag.com/2012/07/ibm-smarter-analytics-signature- solutions/) aims to highlight IBM's more noteworthy portfolio of products in this area, as well as intellectual property that is developed around these products by its services department.

Both self-developed products and products purchased from outside

Can IBM work with Open-source technology from outside, even if these open source technologies compete with some of its own products? Of course. IBM can work in collaboration with R (which is actually a competitor for SPSS), and it has a partnership with Cloudera (its Hadoop distribution competes with IBM's own Open-source version of the Hadoop release) and can use Mahout, Mahout is a stre piece that runs on Hadoop. Of course, there are Linux, and IBM has been strategically using this open-source operating system for years.

IBM also has a number of in-house self-developed products in the field of analysis. Infosphere streams and Infosphere biginsights are two examples of a complex event processing (CEP) solution, which is IBM's own Hadoop release. A noteworthy aspect of the Biginsights Hadoop release is its integration with IBM's DB2 relational database management system, which is one of the most important database management systems in the industry. DB2 database management System. Although relational data may not be big data, previous achievements have laid the groundwork for succeeding. Knowing how to actually process the data establishes the platform and the ability to analyze the data later.

This also involves hardware. IBM is almost equal to the mainframe and the huge backend systems running on mainframes, gathering data for decades. Dealing with a system that has this kind of workload for so long makes big data a concrete concept in IBM's eyes. IBM is not just manufacturing products, ask customers to imagine what kind of data they are dealing with through their products. The imagination is good, but the rich experience of decades is more important.

Focus not only on products, but also on people

With such a large product line in the field of data analysis, it's easy to forget about human resources. But the role of people in IBM's Big data strategy is probably more important than the product. First, IBM has a strong research force, including L.K., which tells me the largest mathematical computing unit in the private sector. This clearly provides IBM with a strong capability in the field of predictive analytics.

The other is IBM in 2002 on the acquisition of PwC's consulting department of the CPA firm. Even before the deal, IBM had a pivotal global service sector, but the acquisition of the Armonk Advisory department presumably made the company of a service-enabled product firm a service company that took full advantage of its parent company's large number of proprietary products.

L.K. and I mentioned another IBM analytics project called Analytics Decision Management (HTTP://WWW-142.IBM.COM/SOFTWARE/PRODUCTS/US/EN/CATEGORY/SWQ60), The project focuses on embedding analytics into business applications, rather than forcing front-line staff to delve into proprietary, isolated analytical applications to analyze and access valuable information. The project, for example, allows call-center employees to understand what products are being marketed for certain callers, and what might be the outcome after marketing. Users of these applications are not even aware that they are using analytical techniques because they are embedded in the operational workflow. It is clear that IBM combined its own experience in scientific research and service delivery to strengthen the ability to meet demand in such a front-line staff scenario.

The conversation with L.K. really opened my eyes. Over the years, I've been focusing on IBM's manufacturing products and acquisitions, and I understand that it's passionate about big data and analytics. But I don't have this in my head in tandem. IBM is in a unique position to do things that competitors cannot do in large data areas.

Blue giants are bad.

At IBM, the following comment is also a bit humiliating: how do other technology companies, especially start-ups, want to build a similar data-analysis empire? How does IBM manage so many disparate products, technologies, consulting teams, and acquisitions of companies? After all, the vast majority of empires eventually fade.

In my opinion, IBM needs to integrate its product line while releasing the new version of the product. In terms of business intelligence, I've seen this happen, and IBM needs to build on it. At the same time, small start-ups don't have to worry about managing so many different components; they are essential to opening up and promoting innovative technologies and markets around these technologies. The Big data is a testament to this.

In the end, however, the industry needs to be integrated. Large data areas are maturing, and more enterprise-class software companies will enter the field, and they will buy up some start-ups, and consolidation will follow. Emerging companies have shown us the importance of idealism and new territory, and IBM's situation has shown us the importance of connecting large data with enterprise-class software and mainstream service organizations. It also shows the significant benefits of embedding analytics into the seemingly mundane business software.

Cutting-edge innovation is critical, but the value of being integrated into mainstream tools, products and companies is fully reflected.

(Responsible editor: Lu Guang)