The technology for voice recognition has steadily progressed so that voice mail, telephone directories, menu scripts in phone services can "hear" and "understand" voice responses. But these spoken commands have considerable limits.
In "Lost in Translation," The Atlantic Monthly, December 1998, Stephen Budiansky said, "One thing demonstrated by all this effort is that language is far more complex than even linguists ever imagined." Crucially, English has a characteristic for determining meaning that makes the process complex: unlike other languages wherein one word has one meaning, in English, all words have 2-5 meanings, each. Thus, an assumption has developed that if you know the part-of-speech, you will know the meaning. Or, if you hear someone's inflection or tone of voice, you will thus know the part of speech, and thereby you will know the meaning. In either case, you will know the part-of-speech, so, according to the assumption, you will then know the meaning.
theExact Word's® technology captures meaning the other way around.
- I ordered minute furniture for her doll house.
- He ran in the door at the last minute.
Maybe you "just know" immediately which "minute" is which. And "of course" you know which is an adjective and which is a noun. But you knew the part-of-speech first, instantly and probably subconsciously in order to know which meaning fits. Your mind may say, essentially, "Oh, if minute is in that adjective 'position,' minute must mean 'little.' " And then, subconsciously, and very rapidly, you will think of the characteristics or attributes of "little." By that list of attributes, you will know that minute in one sentence means small and in another sentence means a unit of time. Of course, you might not have thought consciously about part-of-speech at all. But, to know the meaning at all, you do notice that "position" of the word. The position, the fixed word-order, determines the meaning.
In the computer world, computer programming has emulated the mental process of listing the attributes or characteristics. By listing those descriptive attributes, the program can then distinguish between the two possible meanings. But not either quickly or accurately enough.
theExact Word's® technology lets the computer know which meaning is which first, without knowing the definitions, meanings, semantic attributes, indexed categories, or part-of-speech.
Actually, the meanings in English come from word order. theExact Word® has written the algorithms for word order, how it creates meaning, how parts-of-speech follow word order. Thus, we immediately identify the part-of-speech and, if necessary, the meaning attached to it. Thus, tying natural-language programming to the state-of-the-art in voice recognition and activation can move to a fifth generation speed and accuracy, without semantic attributes.
Go to The
Heart of the Matter--Voice Recognition
Go to the Brainy Read--Voice
Recognition