I had an idea in this vein for what you might do if you were Google. Design your text box widget to capture *all the keystrokes* the user used to assemble the final search string, all the erasures and so on. This would give you a rich corpus of common typos. On Thu, May 7, 2015 at 4:15 PM, Fred Lunnon <fred.lunnon@gmail.com> wrote:
Search engines are now eerily good at correcting typos in their input. I don't know what algorithms are employed to achieve this, but suspect their strategy may have some relevance to this problem. WFL
On 5/7/15, Allan Wechsler <acwacw@gmail.com> wrote:
I think it's time we probed for exactly what you (Dave) mean by "similar". If your criterion for similarity is stricter, it's more likely we'll be able to come up with a speedup over the obvious pairwise comparison.
On Thu, May 7, 2015 at 3:44 PM, Dave Dyer <ddyer@real-me.net> wrote:
Oops, my braino - your number is correct, and not such a big number by today's standards as you pointed out.
The original question remains - is there a better that brute force way to approach it? In actuality, 20,000 is just the number I have on hand, the actual number is much larger.
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com https://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com https://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com https://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun