EMERGING TRENDS FOR THE UPCOMING YEAR
by Ed Colet

Emerging trends for the upcoming year.

As time marches forward in this new millennium, the New Year is always a good opportunity to look back at past trends and link them to future expectations. In this column, I look back at trends in our interactions with technology, and where we might expect to see these trends taking us in the coming years. A noticeable trend is that end-users of technology have expanded from the technically knowledgeable to now include many non-technical end-users. In the more specific field of data mining, this same theme is occurring as well. With this in mind, what might we expect to see next?

Growing the End-User Population:

The increase in the population of people relying upon technology can be attributed in part to the development of the graphical user interface (GUI). Early on, the only way to interact with a computer and use its applications was to know the correct "magical incantations" to type into the command line. For most people, a blinking cursor patiently awaiting instructions was uncommunicative and unfriendly, and limited the user population to technically skilled users.

The GUI changed that significantly. A GUI enabled people to "handle" the applications and "directly manipulate" the information that they wanted to work with. Computers became more user-friendly, and actually more enjoyable to use. A significant development was the advent of the Web-browser that enabled people to not only handle information stored locally, but also to handle information dispersed across the World Wide Web. Applications that used a browser-metaphor emerged to capitalize on users' familiarity with the interface. The Web-browser interface also eliminated obstacles due to different operating systems. It's thought that if you could surf the Web, you could use the application, and the underlying OS/platform doesn't matter (thanks to the Java programming language).

Today, the metaphor of a desktop with files and folders may not adequately reflect the way people currently think of and use information. Some people feel that the Web-browser metaphor may also require fundamental rethinking. Apple, Microsoft, and others are currently re-thinking up new metaphors for an effective user interface. Also, while the browser afforded rich capabilities for displaying information, it does not extend well to other applications that people rely on for information such as hand held computers, cell phones, and other mobile devices.

In the context of data mining technologies, end-users of data mining applications are approaching the same point that end-users in general have reached. But there are still many data mining applications that require advanced knowledge of mathematical models in order for them to be useful, and whose mode of interaction with the user is akin to typing in the right commands (as equation parameters) on a command line (as program scripts). It is only recently that a non-technical end-user, such as a business manager can use such applications effectively (by "non technical", I mean not having an advanced degrees in mathematical analysis). There's growing interest in using data mining on mobile devices as well, which would require careful thought about what an effective interface for data mining on a mobile device should look like.

Broadening the developer population:

Despite the benefits afforded by a GUI for the end-user, the person's interactions are constrained by what an application developer decided about how and when information should be used. And most application development tools are still the purview of the technically skilled. But this too seems to be changing. For example, knowledge of HTML was once required in order to create a Web page. With the development of Web authoring tools, it is now possible for the non-technical end user to create an entire Web site, not merely a page. And in terms of databases, a database programmer was once required in order to build a database application. Today, with the help of wizards, it is now possible for a non-technical end user to create a desktop database application -- complete with tables, queries to run, and reports to generate. All without requiring much knowledge of SQL or programming.

But in the context of data mining, it remains very difficult for a non-technical end user to develop a data mining application. While development tools have made programming simpler through the use of an integrated development environment (IDE) that may automatically generate program statements, or import the appropriate classes and methods, an IDE is still not usable by a non-programmer.

Can we ever expect that a non-programmer such as a business manager will be able to develop a sophisticated enterprise-wide data mining application? The answer is not only a resounding "yes", but also, that the non-technical business manager is precisely the person that should be building his data mining applications. It is the business manager that best knows how he/she wants to interact with the information; coupled with intimate knowledge of the business domain, this person is also best able to interpret results. A tool available from us at Virtual Gold, called AppBuilder has been designed with exactly this purpose in mind, and is already currently available.

While the time is right to take stock of the past and form expectations about the future, it's fairly certain that the pervasiveness of technology will continue. As more of us become increasingly dependent on computer applications the expected population of end-users as well as developers are likely to include greater numbers of "non-technical" end users. Providing these users with capabilities for them to develop their own applications to best suit their needs is likely to be an emerging direction in this new year.

Ed Colet is the Acting Director of Research at Virtual Gold Inc., responsible for developing analytical methods for data mining and for investigating human factors and usability issues of business intelligence systems. At present, he is in the final stage of completing a doctoral dissertation in the Cognition and Perception program at New York University's Department of Psychology. Ed has also worked for IBM Research at the T.J. Watson Research Center. At IBM, Ed was a member of the group that developed Advanced Scout, the data mining application for NBA teams. His research interests focus on statistical methods and human factors.

For more information, see www.virtualgold.com.