Mobile Edition
Print
Get the Mag
Weekly UpdatesIntroduction: Vern McCorkle
This is the fourth in a series of Executive Forum sessions on the Knowledge Economy. This time, we examined the use of mega databases and addressed theories such as: What technology has built mega databases? What techniques organize and manipulate them? How have we used huge amounts of data to improve the way we live and work? What are the skills we need to operate in this data-hungry environment?
Thomas R. Case, Dean, UAA College of Business and Public Policy
Knowledge is not an either/or choice between a resource-based economy and a knowledge-based economy. Technology and information processing enable resource management. In my lifetime, we have moved from mechanical technology, like the old Underwood typewriter, through the IBM Selectric that had a correction feature, to systems that translate voice into written text. The rate of change is mind numbing. Innovation opens a wider capacity for data collection and management. The mega data capacity uses the power of information management to expedite solving business problems and offers a wide array of ways to add value to the system.
Dr. Mini Yen, Ph.D UAA CBPP Basic Issues of Database Design
"Garbage in = garbage out" is the most important concept to learn. Basic design of a database is critical for the successful operation of the entire system. Two kinds of database management systems address the design of the database, manage and administer the data and establish appropriate security systems.
First, online transaction databases are for day-to-day transactions in real time. It is important to adapt old-fashioned flat files into electronic form with the normalization and authentication of the data. Failing to build relationships among the functions in a database will yield confusion and inaccuracy. Every data point must have a unique identifier to make a record meaningful and to avoid orphan data that cannot be traced back to its origins. All the data in each table must relate to only one concept. This keeps the information from being confused.
A second kind is the data warehouse function. It is more appropriate for storing data of historical value, time series information and collections of information over time. This kind of database has a focus on data validation and retrieval for analytical use. Its power comes from speed of data retrieval and the volume of data handled. This kind of data design supports executives and decision-makers and is often used for building models and supporting analysis of facts. It usually has a time dimension in addition to other specific data, such as product or customer information and geographic information.
Stuart Ferguson, Ph.D., Alaska Federal Health Care Access Network
This example from June 2003 captures the importance of tele-health in Alaska. A woman in Kotzebue had an eight-week ectopic pregnancy that ruptured. There had been a gunshot emergency the day before that drew down the blood supply in the clinic, and the woman was bleeding extensively. A doctor in Anchorage had a live feed set up in Kotzebue and talked the health aide through the treatment procedure. Telehealth saved this woman's life.
The telehealth project allows specialists in Anchorage to serve rural areas in real time. The project serves a diverse statewide population of Alaska Natives, military and veterans whose needs defined the telehealth system. We faced problems of variable power systems in the villages, modest to absent computer skills, and the villages could not easily share data. We had to deal with a satellite six-second delay in relay of data, and there were fidelity of data issues. We deployed 300 hardware systems, developed a software system and a network of servers that communicate every five seconds. Now, data can be pushed from 700 providers throughout the state and served 23,000 cases last year. We see 50 percent growth in use per year. More than 6,000 cases avoided travel, at savings of about $600 to $800 per patient.
Telehealth uses mega databases to accumulate and manage the packets of data in the form of images and readings. Telehealth systems use expertise in capturing and moving data to where it needs to be, but are not yet experts in pushing data into the electronic health records for a patient. The data should be able to move to wherever it is needed as a patient moves from one system to another. Telehealth is in its infancy. It affects telecommunications systems due to need for T1 lines. (A 1.544 Mbps point-to-point dedicated, digital circuit provided by the telephone companies.) Telepharmacy, tele-consulting to other Native organizations and other countries all offer business opportunities we are only beginning to explore. In Alaska, we are becoming the source for documenting the use of telehealth systems. We are now able to export this expertise and generate revenue for sustained activity
Scott Monnig, Hanager, Information Technology Provisioning for ACS
The physical environment is structured from the bottom up and we look at customers from the same perspective. From the largest data down, we see call volume of 19.25 million transactions per week--1 billion calls annually. Our data system maintains those records in retrievable form for seven years for law enforcement and regulatory records. We are moving into overlap areas, which we have not had to address before, such as marketing issues--What market segmenting? Can we identify? What are the transitional spaces for growth? We need to look at the data in new ways, and add data from other sources we do not generate and store internally. Integration with telehealth is one example of a difficult interface for data management as well as other business issues.
The future is in moving from data tracking to warehousing data and using it for analysis and strategic planning. Change is rapid and continual. The data system is the enabler for moving the entire industry forward. Computers really only automate repetitive analysis tasks; you still need the brain to attach value and to shape the questions for analysis. We need people with embedded curiosity and continuous learning. They must know how to learn on an ongoing basis over their entire professional life. Pattern recognition, abstraction, relationship identification and communication are the essential skills.
Dede Schwartz, Geodata Manager, ConocoPhillips Geodata Applications of Mega Databases
Data management at ARCO in the 1980s involved colored pencils on the wall. Today we have 35 terabytes of geophysical data. The challenge is to select the right set of geophysical information from an overwhelming array, to keep the data moving and maintain quality as it moves between systems. Data collection is a valuable resource, and we do share it externally with partners, regulators, service providers and sometimes competitors.
We also share resources data electronically with our partners or potential partners, especially those interested in exploring or developing areas we are not interested in. To be useful, the data bank is enormous. For example 30 of 35 terabytes is raw seismic data. Exploration is very data hungry. Real-time data feeding from drilling sites allows trouble shooting and service without deploying personnel to dangerous or remote locations. It is also used for world drilling planning. Using seismic data to view geologic information over a global deployment can direct drillers in the field from the remote drilling centers. Cameras installed in the drilling platforms help this process, which generated human interface problems. Documenting the origin and validity of the data sets is an important aspect of this kind of system. Virtual reality can be manufactured in science fiction mode, and it is sometimes difficult to separate simulations from valid real-time data. People think: "If it is on the computer, it is automatically right." Much of the information available on the Internet is not properly vetted for accuracy and validity!
DISCUSSION:
Vern McCorkle: We began this series with the intention of connnecting the academic community with the business community two years ago now. Here we see a good example of how the university can be involved in applications. KISS now means Keep It Simple Smarty, because it is much harder to make things simpler than it is to fall back into "garbage mode."
Cheryl Wright: Several speakers touched on how to make databases useful. Complex systems help to ask the questions that allow data to be meaningfully applied. Move toward optimization of organizations and systems. How do you know what part of your data system can be valuable in solving an instant problem? The university is a place where people in business can come together to apply this kind of information analysis to solving problems.
Dede Schwartz: The CBPP computer internship program with ConocoPhillips has been tremendously successful. We sometimes shape the project to the skill strengths of the student, but usually, they adapt their skills to the task. We hire many of them, and they get promoted out to the industry.
Thomas R. Case: Decision-making involves the brain, not the technology. It is easy to put data into a database, but it takes a well-designed system to get information out! We need to keep our focus on the K to 12 students. Kids at age five are facile with computer literacy. They think to use it automatically. The future holds exponential expansion of opportunities with data as a tool for creative minds.
Julie Kitka: The knowledge-based economy series has been very helpful. We will have a follow-up conference to look at the economic implications.
Participants:
Vera McCorkle, Publisher, Alaska Business Monthly
Thomas R. Case, Dean, UAA College of Business and Public Policy
FEED