Advisor Today > > August 1999 Digital Agent: Improving Input Capacity

Digital Agent: Improving Input Capacity

By Edwin P. Morrow, CLU, ChFC, CFP, RFC

You have always believed the two essential peripherals required to operate your computer are the monitor and keyboard. The monitor allows you to know what the computer is doing, and the keyboard is required to enter data and instructions. You will always need those two elements, whether or not you have a scanner, external drives, printers, or a modem, right?

Wrong! The most rapidly expanding technology for the personal business computer is the incorporation of the auditory senses. All new computers now come with a sound card or chips to interpret digital data and convert it into sound waves. The actual output device may be speakers or a headset. However, if you are planning to make significant authoring use of a computer, then you will probably be connecting a headset with a microphone for voice input.

To embrace this new technology is pretty simple: Buy software, install a microphone/headset, and train your computer to accept your voice inflection. Give your fingers a rest!

There are four leading software providers: Dragon Systems, IBM ViaVoice, L&H Kurzweil and Phillips FreeSpeech. In a live demonstration at the most recent Comdex convention, the audience saw several products demonstrated from the main platform, and voted the Dragon product to be the leader.

There are four general categories to consider when evaluating speech recognition software: setup and training; editing and formatting; application integration, and command-and-control. For example, Dragon has a "new user wizard" that guides you through training the vocabulary to your speech patterns. Editing ease is critical, since you can initially expect only 91 percent accuracy of input (which may rival your typing quality). You can expect most programs to integrate with the Microsoft suite of products, and a few with Corel.

In its exhaustive tests, PC Magazine reported (11/98) "To gauge accuracy, we used several different tests. In the first test, we asked a team of six testers with no dictation experience to dictate a series of documents into each program. NaturallySpeaking came out on top. On a 1,000-word business document, NaturallySpeaking will make 38 fewer word errors than its closest competitor, L&H Voice Xpress Plus."

Personally, I have only tested L&H Kurzweil and Dragon, and Dragon was clearly superior, although L&H had some nice command features. All of these vendors are in a race to capture the most efficient method of entering data into computers, so you can expect fairly frequent product updates and good support.

Dragon comes with its own high-quality headset and excellent installation instructions. No matter which software you choose, you will need a fairly powerful computer, since you will normally be running at least two programs at once, the voice recognition and a word processor or spreadsheet. Hardware requirements: 32 MB RAM minimum, 84 MB recommended; 32 MB of free hard drive space, but obviously more will be needed to store documents. You also need a CPU processor of at least 133 MHz, but 200+ recommended; and a sound card that is compatible.

If you don't get the best reception to your voice, you might try a high performance headset mike from Andrea Electronics Corp. (ANC-700/

750; list $149.95). Another alternative is the Phillips Electronics SpeechMike Pro for $150 that includes a handheld microphone, trackball and speaker with command and control buttons.

You may comparison-shop for a speech recognition program but I suggest you talk to a knowledgeable person to be sure you are ordering the preferred model. Dragon has several less expensive units, but the serious business user will pay about $500 for the preferred version with greater vocabulary capacity and the latest features.

Can you use a speech recognition with your database or graphic software? The answer is a very qualified, "Maybe now, but most likely later." However, first you will have to become accustomed to this entry method and get your system trained to your voice.

It may be possible to operate both voice and database/graphic programs simultaneously and reduce your input time—but you will have to do some testing. Clearly voice delivery is faster and more accurate than nearly any human can type.

Edwin P. Morrow, CLU, ChFC, CFP, RFC, is president of Financial Planning Consultants, a firm based in Middletown, Ohio. He is a consultant to financial advisers in practice management and computerization and the author of seven software programs including the Text Library System. Readers may contact him at: Box 42430, Middletown, OH 45042, telephone (513) 424-1656, or e-mail edm@financialsoftware.com.

Top | Back | E-mail to a Friend | Print

If you are not receiving your magazine or need to change your address,
please contact membersupport@naifa.org.

For comments on articles on this website contact mleyes@naifa.org.