In spite of the plethora of intelligent, talking computers familiar from the movies, we all realize that our real-world computers – from the most monstrous machine in a data center to the sleekest smartphone – only comprehend the language of zeroes and ones. And yet, technology exists today that understands much of the meaning of human language.
Such technologies and the science behind them are known as semantics. And in a world where a computer can defeat the best human players at a game of Jeopardy!, who are we to say whether the computers really understand us, or are simply getting better at maintaining a pretense of such understanding?
The problem with semantics is that it comes down to the human understanding of language in all its vagueness, ambiguity, and context sensitivity. In many cases, efforts to standardize meanings strip out vagueness and ambiguity from data when the goal to foster precise, unambiguous, and consistent communication between systems.
The problem with this approach, however, is that human communication is inherently vague, ambiguous, and relative. The way humans understand the world, the way we think, and the way we put our thoughts into language require both vagueness and ambiguity. Without them, we lose important aspects of meaning.
As a result, other efforts at building semantics into software focus on the harder problem of dealing with inherently context sensitive and dynamic human meaning, in order to establish automated conversations with individuals. Either approach, or perhaps a combination, may form a critical part of your Digital Transformation initiative – if not now, than in the near future, as semantic technologies inevitably improve. Understanding the power and limitations of semantic technology, therefore, is an important arrow in the arsenal of any Digital professional.
Read the entire article at http://www.forbes.com/sites/jasonbloomberg/2014/08/25/teaching-computers-to-understand-meaning-from-banking-to-healthcare/.
Image credit: Peyri Herrera
