h a l f b a k e r yLoading tagline ....
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
If you're writing a document and type the word "compliant" 10 times, one of those times you might type "complaint" by mistake, and the spell-checker won't catch it because it is a real word.
Similarly, if you work in customer service, you might write the word "complaint" 10 times in a document but
write "compliant" by mistake once, and you'll never notice it and it will look stupid.
Enter the Subject-Sensing Spell Checker. This feature searches for words that occur often* in a document, then scans the document for words similar to these base words. (Derived forms must be ignored during this search. That is, even if "window" occurs 20 times and "windows" only once, that is not to be flagged. It is likely to be correct.)
* "Often" means "relative to the normal frequency of the word in the English language." The computer checks each word against a list of the 10,000 most common English words. So in the above example, "complaint" is obviously significant since it normally occurs every 1 million words and you've used it 10 times in a 900-word document.
Hopefully, with proper filtering, no one will be embarrassed by a Public Services document wherein a building is referred to as the "pubic library".
Nor will eager porn readers become bewildered during a sex scene by the reference to someone's "public hair".
[link]
|
|
surely pornography makes pubic hair public. |
|
|
I really like this idea, but I'd imagine that the program would require a considerable of training before it was of any real use (as with those programs hat type worlds as you speak them - hose thongs drivel me inane). |
|
|
'Public Hair' was a Hairdressers shop in Nottingham. A Snip at the Price!. Back on topic, I could see this patch being as annoying as the original problem. It must be very difficult to judge context and subject at the same time (particularly for more flamboyant writers). I have also noticed that my T9 dictionary tries to do this by trying to finish the phrase. Unfortunately it uses the last one that you typed. |
|
|
I don't much like using the same word ten times in a document. |
|
|
-1- People who mostly mean one word but might accidentally type a similar word by mistake when they still mean the first one. |
|
|
-2- People who don't understand that the word they're using isn't really the word they mean, such as when describing a double-axel trailer (is that a trailer that rotates 2.5 times when jumping?). |
|
|
It might be useful to have automatic pop-up "sticky notes" to indicate which words in a document were commonly confused, and have a training function so that someone who knew the difference between, e.g., "infer" and "imply", wouldn't be continuously bothered. |
|
|
I have no intention to cover People #2 in this program. This is only a spell-checker. |
|
|
Computerized grammar checking is horrible; this program does not attempt it; you still need a proofreader for that. But, if you *meant* to type "axel" and you did so 10 times, but once typed "axle", then my program will CORRECTLY flag the word "axle" and ask if you'd like that changed to "axel", too. |
|
|
Sounds good to me. I once called the
chief financial officer of my organisation
"spectacularly incompetent" when I
meant to say "somewhat mistaken". My
email's spell-checker let me down badly
on that occassion. |
|
|
...You had already called him incontinent five times, but you accidentally wrote "incompetent" when it really mattered. Bummer; I think my product will help you. |
|
| |