 |
 |
 |
|
|
A "false positive" is when a word or series of letters gets blanked out because
a filter has interpreted it as an offensive word. For example, take the
word "scrapbook". Many filters would blank out the
letters "crap" producing an unacceptable and annoying "s****book"
in your blog post.
We checked these Great Literary Works
Disinfected Text has been tested against several great literary works to
uncover many common "false positive" scenarios. Catching all of them
would be nearly impossible but the most common have built-in countermeasures to
prevent them from occurring.
Below is the current test set we use as a base to check for "false
positives":
|
|
 |
|
|
|
| Disinfected Text - "False
Positive" test cases |
False Positives |
| Title |
Author |
Words |
Before optimization |
After |
| The Holy King James Bible
|
God |
773,692 |
* |
* |
| Frankenstein
|
Mary Shelley |
69,701 |
* |
* |
| The Count of Monte Cristo
|
Alexandre Dumas |
465,892 |
* |
* |
| Lady Chatterley's Lover |
D. H. Lawrence |
118,460 |
* |
* |
| Robinson Crusoe |
Daniel Defoe |
121,650 |
* |
* |
| The Tell-Tale Heart |
Edgar Allan Poe |
2,172 |
* |
* |
| Tarzan of the Apes |
Edgar Rice Burroughs |
85,784 |
* |
* |
| Jack And Jill |
Louisa May Alcott |
70,878 |
* |
* |
| Alice's Adventures in Wonderland |
Lewis Carroll |
26,947 |
* |
* |
| War And Peace (books 1 - 8) |
Leo Tolstoy |
273,021 |
* |
* |
| The Adventures of Huckleberry Finn |
Mark Twain |
112,605 |
* |
* |
| A Christmas Carol |
Charles Dickens |
28,614 |
* |
* |
| The Three Musketeers |
Alexandre Dumas |
230,188 |
* |
* |
| Dracula |
Bram Stoker |
163,136 |
* |
* |
| The Wonderful Wizard of Oz |
L Frank Baum |
40,689 |
* |
* |
| The Land That Time Forgot |
Edgar Rice Burroughs |
37,308 |
* |
* |
| The Hound of the Baskervilles |
Conan Doyle |
59,626 |
* |
* |
| Tale of Two Cities (books 1 - 3) |
Charles Dickens |
137,050 |
* |
* |
| The Adventures of Peter Pan |
J. M. Barrie |
47,957 |
* |
* |
| The Adventures Of Tom Sawyer |
Mark Twain |
66,085 |
* |
* |
| Around the World in Eighty Days |
Jules Verne |
63,215 |
* |
* |
|
* Optimization is complete, numbers will be posted by May 2006
(human intervention required to discern false positives
from actual offenses)
|
|
|
|
|
