FREE DM Review Site Registration!
Sign-up today and access DM Review on the Web!

Your FREE registration entitles you to:
FREE email newsletters

FREE access to all DM Review content

FREE access to web seminars, resource portals, our white paper library and more!

   

Data Mining Tools: Which One is Best for CRM? Part 2

Editor's note: Part 1 of this series appeared in DM Direct on January 24, 2006.

The Knowledge Discovery Process

One way to refer to this wider context of data mining is to include data mining as part of the broader process of knowledge discovery (KD). In addition to data mining activities, KD may also contain some elements that define data extraction from source systems, customer householding systems and actionable systems that receive input from (driven by) data mining analyses. These actionable systems might include business intelligence systems to create reports for management, marketing campaigns and marketing results tracking systems. While the availability of all of these KD systems are necessary to assure the success of the data mining system, the many tasks that compose them are beyond the scope of this analysis.

Once the knowledge has been discovered, it must be transformed into modes that can serve as appropriate inputs to various business processes that can generate increased profitability.

Business Process Management

Business process management is the management of complex interactions between people, applications and technologies in a business designed to create customer value. Business process management takes knowledge discovered in analytical operations, interprets it in terms of current business objectives, and composes it into actions to serve current business objectives. Business process management includes quality programs, such as Total Quality Management and Six Sigma, as well as development programs, such as Function Point Analysis.

Knowledge Management

An even broader context for data mining beyond business process management is the higher-level process of knowledge management (KM). Mentzas loosely defines this process to include the broad collection of organizational practices and approaches involving generating, capturing and sharing knowledge that is relevant to the organization's business. Mentzas, et al. distinguish two common approaches taken in KM are the product approach and the process approach.

The product-centric approach: This approach focuses on documents, storage, case histories and solution templates. The authors point out that adopting a product-centric approach treats knowledge apart from the people who create (or discover) it, and the people that use it.

The process-centric approach: The process approach takes a more holistic approach to knowledge management by emphasizing the environment in which knowledge is generated and distributed. The authors view the process-centric approach as a social communication process. This means that knowledge is centered on those who discovered it, and distribution of knowledge occurs via person-to-person contacts. The process develops knowledge communities as self-organizing groups that participate in communication as a natural and evolving function.

These distinctions in KM are not novel; they have roots in the dichotomy of thinking that characterizes human thinking. Reflections of these roots extend in literature all the way back to Plato and Aristotle. I discussed these reflections in their relationship to customer behavior modeling in a previous article.3

Aristotle believed that the true being of things could be discerned by what the eye could see and the hand could touch. He believed that the highest level of intellectual activity was the detailed study of the tangible world around us. Only in that way could we understand reality. Based on this approach to truth, Aristotle was led to believe that you could break down a complex system into pieces, describe the pieces in detail, put the pieces together and understand the whole. For Aristotle, the whole was equal to the sum of the parts. Aristotle viewed this nature of the whole as very machine-like. Product-oriented knowledge management is very Aristotelian in its approach to truth.

For Plato, Aristotle's teacher for 20 years, the only thing that had lasting existence was an idea! He believed that the most important things in human existence were beyond what the eye could see and the hand could touch. Plato believed that the influence of ideas transcended the world of tangible things that commanded so much of Aristotle's interest. For Plato, the whole of the essence of being was greater than the sum of its tangible parts. Process-oriented knowledge management is very Platonic in its approach to truth.

Knowledge and Business Ecosystems

We must not permit our thinking to become polarized in our pursuit of knowledge. We must view our business world through the binocular lenses of Aristotle and Plato combined.4 This is what is done now in modern ecosystem analysis.5 I was trained first in the scientific neo-Platonism of plant sociology. Later, I was trained in the scientific Aristotelianism of physiological plant ecology. But, it was Daniel Botkin (my mentor for seven years) who showed me how to put the pieces together to model forest ecosystems. Botkin's approach to forest modeling is very analogous to Mentzas' approach to KM.

The analogy between the approach to truth in ecosystem analysis and business was first suggested by Claudia Imhoff and Ryan Sousa.6 These articles were expanded into a book with Bill Inmon.7 Mentzas' unified KM approach of calls for "balanced fusion" of these two approaches to define what Mentzas calls "the Know-Net approach." The Know-Net is defined by Mentzas as a corporate information structure that enables individual, team, and entire organizations collectively and systematically to create, share and apply corporate knowledge assets to better achieve organization efficiency, responsiveness, competency and innovation.

The focus of the Know-Net is on knowledge as the critical strategic resource in an organization. This focus is almost exactly in parallel with the focus on energy flows in ecosystem analysis. In a very real sense, knowledge is potential business energy. The challenge for IT managers is to design systems that can take the knowledge discovered in data mining activities (potential business energy) and convert it to business kinetic energy to get things done throughout the organization that enhance corporate profitability. This overall process must integrate the erstwhile separate processes of data mining, knowledge discovery, business process management, and knowledge management. The best data mining tools, therefore, are those that help the most to make this happen.

Data Mining in Business Ecosystems

In order for data mining tools to serve well in the Know-Net, they must facilitate the business processes that precede the data mining project (e.g., data storage and retrieval, data extraction, integration), as well as those that follow the project (model export, model deployment, model validation). Also, data mining tools must facilitate the transmission of knowledge flow into knowledge management processes beyond the data mining project. Such a holistic view of data mining places it in business ecosystems in a role analogous to that of photosynthesis in natural ecosystems. Radiant energy is absorbed by plants and transformed into chemical energy structures (sugars), which serve to carry energy into higher levels of the food chains. These chemical energy structures must be compatible with the input requirements of higher organismic processes. This organismic view stands in stark contrast to the traditional view of companies as machines grinding out production. Data mining tools must be viewed in this organismic context, or they will often fail to achieve the business goals that conceived them.

To be able to realize this goal, data mining tools should:

  • Facilitate data interchange between the data mining activities and the rest of the business enterprise (common data structures):
    • SPSS data format,
    • Statistical data format, and
    • SAS dataset.
  • Facilitate model interchange within the business enterprise (PMML, SAS code).
  • Interface with common BI tools.
  • ODBC and Call-level data interfaces.
  • In-database data mining capabilities.
  • Facilitate inclusion of user-defined nodes or plug-ins in increase connectivity with other systems.

Eventually, data mining tools must permit development of a Know-Net, which passes information around in an enterprise to enable collaborative decision-making. Bill Gates calls this the "digital nervous system" of a company. Only in this way, can an organism develop multi-component synergies like hand-eye coordination. Likewise, the business ecosystem can only develop when the parts of an enterprise work together to do "business at the speed of thought."

References:

  1. U. Fayyad, G. Piatetsky-Shapiro and P. Smyth. Advances in Knowledge Discovery and Data Mining, Fayyad et al (Eds.) MIT Press, 1996.
  2. Gregor B. Mentzas, et al. Knowledge Asset Management. London: Springer, 2003.
  3. Bob Nisbet. "How to Choose a Data Mining Suite." DM Direct Special Report. 24 March 2004. Nisbet, 2004.
  4. Daniel Botkin. Discordant Harmony: A New Ecology for the Twenty-first Century. London: Oxford, 1990.
  5. Claudia Imhoff and Ryan Sousa. "Corporate Information System in Action: Birth of the Information Ecosystem." DM Review, 1997.
  6. Bill Inmon. The Corporate Information Factory. New York: John Wiley, 1998.
  7. Bill Gates. Business @ the Speed of Thought. New York: Warner Books, 1999.

Robert A. Nisbet, Ph.D., is an independent data mining consultant with over 35 years experience in analysis and modeling in science and business. You can contact him at Bob@rnisbet.com or (805) 685-0053.

For more information on related topics, visit the following channels:



Industry Vendors