| Natural Language Technology |
| A Bridge to the Information You Want |
Responding to ad-hoc or even recurring queries of data warehouse information usually requires the intervention and considerable resources of an
overworked IT department. For all practical purposes, the data is there but cannot be retrieved in a timely manner. New methodologies and
approaches are emerging that offer a leap in capability that directly addresses these kinds of shortcomings.
Natural language processing can be thought of as a way to communicate with your system in the same way that you communicate with another person. This involves
"understanding" the concepts that are intended by the words that are typed or spoken, a very difficult task. It's ironic that natural language, the system
that is easiest for humans to learn and use, is hardest for a computer to master.
However, software tools exist that do a very good job of bridging the gap between an English sentence and a database of information from which an answer is desired.
The problem is different from that which an Internet search engine solves, where a set of "close matches" is returned from a give search phrase. A true natural language system should yield
the exact answer desired, not just a bunch of close guesses.
|
| We Can Help |
|
The Intellection Group has developed, as part of our core framework, a methodology to incorporate an effective natural language processing engine into business
applications, even web-based business applications. While it may not be feasible in every instance, our experience shows that a semantic model can be effectively
implemented on top of a well-structured data schema. This allows a regular English sentence to create a direct path to your answers without having to know how to formulate
a SQL query. A very powerful tool that is underutilized in business today.
|
| |
|
|
|
