home overview techniques technology resources about


Future Applications

Imagine a computer that could communicate through speech or text just as a human could but with the added knowledge and information resources of a computer. Many scientists dream of such a “perfect” natural language processing machine, but there is some consensus that such a machine would require many other artificial intelligence technologies to become significantly more advanced in order to operate. Thus, most current research is aimed at furthering existing technologies and expanding their scope.

Improving accuracy in natural language processing is a constant goal, as even a small percent error results in a large number of errors due to the vast amount of data that is often processed. Consequently, research is invested in finding ways to reduce to reduce error rates while keeping the time necessary to compute algorithms reasonable. If error rates could be minimized, one technology that would benefit is speech recognition. Lower error rates would make this technology more useful to the general public as well as improving the experiences of those with disabilities who already use the technology.

Another attainable goal is expanding the ways in which algorithms analyze text for processing. As computing power expands, programs can make use of the extra resources to gain more understanding of context and employ more complex learning algorithms. Such advances would overlap with other fields of artificial intelligence, and natural language processing research could benefit from such a pooling of knowledge and resources. Ultimately, the success of natural language processing may be extremely dependent on the success of other machine learning and knowledge representation technologies.