Volume 14, Issue 5
|
|
Volume 14, Issue 5
|
|
|
|
| Sept/Oct 2000 | |
| Theme: Data Mining, Genetic Algorithms, Modeling and Simulations |
|
To Volume 14, Issue 4
|
To Volume 14, Issue 6
|
||||||||
 |
 |
||||||||
|
Features
|
| From Data to Insight: The Critical Path to Data Mining, A Short History of Data Mining Lou Agosta explains how to advance from raw data to actionable business insight using scientific methods as implemented in data mining applications. |
| Free-Form Text Data Mining: Integrating Fuzzy Systems, Self-Organizing Neural Nets and Rule-Based Knowledge Bases Earl Cox describes the application of a hybrid approach to finding hidden relationships within large databases revealing deep insight and generating significant financial and commercial results. |
| AI@Work Loan Logic Calculator: Rules-Based Programming for Loan Application Processing; Molecular Data Mining Tools: Advances in HIV Research; Intelligent Investments n the Field of Uncertainty: Analyzing Engineering Projects Using Real Options. |
| Tackling Real-World Environmental Engineering Challenges with Linear Genetic Programming Larry M. Deschaine advocates a unique approach to the challenges of engineering and scientific data mining, control, and process optimization by using fast linear genetic programming technique. |
| Simulation and Modeling: Engineering's Unified Approach Roger G.E. Franks and Jon Paul Van Buskirk investigate the combined use of simulations and artificial intelligence to improve research and development, plant design, retrofitting and operations and ultimately better business decisions. |
| Webbed Feet: A Walk Through PC AI's Updated Web Site Ilana Marks explores the enhancements to PC AI's AI web site. |
| Regulars | ||
| Editorial | ||
| Intelligence Files - Here Yesterday, Gone Today, Back Tomorrow | by David Blanchard | |
| AI and the Net - Filtering the Web | by Mary Kroening | |
| The Book Zone - Data Analysis & Decision Making with Microsoft Excel and Design of Experiments: Statistical Principles of Research and Analysis, Second Edition | by Will Dwinnell | |
| Product Updates ----------------------------> | 26 late breaking product announcements from around the world in the fields of: | |
| Business Forecasting | Decision Support | |
| E-Business Solutions | Expert Systems | |
| Expert System Development Tools | Help Desk | |
| Intelligent Portals | Intelligent Tools | |
| Internet and Web | Languages | |
| Machine Learning | Market Analysis | |
| Modeling and Simulation | Natural Language Reasoning | |
| Object Oriented Development | Training | |
| Announcements | ||
| Product Service Guide - Provides access to information on an entire category of products | ||
| PC AI Blackboard - AI advertisers bulletin board | ||
|
Advertiser List for 14.5
|
||
| AAAI | Exsys Recruiting | Salford Systems |
| Abstract Developers | Franz Inc | Soliloquy |
| AI Developers | Frontier Global Systems | StatSoft |
| Amzi! Inc | Logic Programming Associates | The Center for Data Insight |
| AND Corporation | Megaputer Intelligence | The MathWorks |
| Angoss Software Corporation | NeuroDimension Inc | The Modeling Agency |
| Applied Logic Systems | PC AI Back Issues | Ward Systems Group Inc |
| BioComp | PC AI Banner Ad | WizSoft Inc. |
| California Scientific Software | PC AI Reprints | |
| Cyber Squire | Production Systems Technology | |
| DCI | Prolog Development Center | |
| Decisioneering | QMC | |
| DTSoftware | RML | |
| Exsys | Rule Automation |
| Most of the time when we discuss knowledge in the context of artificial intelligence, we talk about the hard-fought battles of extracting what people know and codifying it into rule-bases or case-bases - or even more mundane - just writing it down in prose. Yet some knowledge has not been learned, remaining hidden in vast relationships between products and services, and the customers who buy them. This knowledge is ripe for exploitation with data mining and business intgelligence tools. Hidden in these rows and columns are information nuggets relative to buying trends and interactions, cross-selling opportunities and areas where one selling techniques may work better than another. When overlaid with external sources, like demographic databases, local trends can be interpreted. | |
| Of course, the technology that identifies these interpretations of context and data is AI in and of itself. The statistical probability and pattern recognition brings us full circle from tools that capture knowledge, to knowledge-based tools that interpret knowledge from structured data. | |
| What none of these technologies can do, however, is bring the derived knowledge back into a human setting where action can be taken. Even if I learn that customer consistencies are an ideal match for a new product, based on their geography, their buying trends and their involvement with web-based communities, I still don't have a plan. The knowledge base can only point out what it infers, not tell the management team how to proceed. | |
| There is a fundamental disconnect between all types of automated interpretation systems and the people who must act upon these interpretations. It is important that readers recognize this because it is an issue dealt with today through the application of a business process. When a system makes a discovery, the analyst concurring with that inference must place the idea into a collaborative structure where it can be acted upon, perhaps even improved. | |
| Some organizations look to their data warehousing operations as the center of their knowledge management initiatives. But data warehousing and the subsequent mining of that data for interfered knowledge is only a starting point. The real knowledge management kicks in when the organization reacts to that new knowledge, embraces it, takes action on it, and nurtures it through the inevitable changes that it will incur once exposed to the light of human intellectual curiosity. | |
| With data mining, the analysts generate a hypothesis from which they attempt to discover underlying data that satisfies that hypothesis, just as a Prolog program searches through its data structures for justification, returning at times with a null result. Organizations do not understand that just asking a question and finding an answer is only the first step along the journey to knowledge. These organizations will likely find that null set staring them in the face when they attempt to justify various technology solutions like data warehousing, data mining, or business process simulation. The technology can only suggest, only help refine, only codify a design. It cannot implement, it cannot improve and it cannot reflect on its discoveries. Those acts remain the purview of people, and it is crucial that technology leaders, when dealing in the realm of knowledge, overtly recognize the necessary inplications of their automation's inferences, and create a culture that can transform the interpretations into business value for customers and stock holders. | |
|
Daniel W. Rasmus |
|
||||
|
|
|
|
|