Re: A Newbie Question - Matching Two Texts

Entity extraction is viable for some applications.

You have to be willing to accept a significant error rate. This is a
show stopper for some situations, but a non-issue in others.

You also have to be willing to put a fair bit of work into tuning and
re-tuning your application. Named entity extractors are currently very
domain specific and need a fair bit of hand-holding. For low to
moderate volume situations, it can even be cheaper to do the task by
hand. Only for high-volume and (error tolerant or high-value)
applications is named entity extraction a very plausible option.

[ comp.ai is moderated. To submit, just post and be patient, or if ]
[ that fails mail your article to <comp-ai@xxxxxxxxxxxxxxxxxx>, and ]
[ ask your news administrator to fix the problems with your system. ]
.