验证码 [英] captchas
问题描述
我在想一种使用类似想法的方法，但要大幅度地增加bwot机器人的能力。
想法是使用认为人类可以使用上下文和模式来理解比以下情况下的计算机更好，
http://www.simplebits.com/notebook/2...sed_wdors.html
所以我的想法是向用户询问一个非常简单的问题，这些问题是使用
拼写错误的单词写成的，这样就几乎不可能了。
计算机识别答案是什么。这个问题可以通过图形方式进行操作，以便它在视觉上更难以像平常一样弄清楚
验证码。
除了改变一个单词中字母的顺序可以故意地改变字母，以便这个单词没有意义，但在上下文中对于人类来说并不困难。甚至可以使用糟糕的语法和一些其他的东西来使它变得更加困难。
我以为可以创建一个非常大的数据库
网站可以使用的不同语言。
你们怎么想？
Jon
I was thinking about a way to use a similar idea but to dramatically
increase the ability to thawrt bots.
The idea is to use the idea that humans can use context and patterns to
understand much better than computers in such cases as the following,
http://www.simplebits.com/notebook/2...sed_wdors.html
So the idea is to ask the user a very simple questions that is written using
misspelled words and such so that it makes it almost impossible for
computers to recognize what the answer is. The question can be graphically
manipulated so that its also visually harder to figure out like a normal
captcha.
Besides changing the order of the letters in a word one can deliberately
change letters so that the word doesn''t make sense but in the context is not
difficult for a human. One can even use bad grammar and a few other things
to make it more difficult.
I was thinking that one could create a very large database of such things in
different languages that sites could use.
What do you guys think?
Jon
推荐答案
5月9日下午1:25，Jon Slaughter < Jon_Slaugh ... @ Hotmail.comwrote：
On May 9, 1:25 pm, "Jon Slaughter" <Jon_Slaugh...@Hotmail.comwrote:
我正在考虑使用类似的想法，但要大幅增加
增加解冻机器人的能力。
这个想法是使用人类可以使用上下文和模式的理念来比在这种情况下的计算机更好地理解以下，
http ：//www.simplebits.com/notebook/2...sed_wdors.html
因此，我们的想法是向用户询问使用的非常简单的问题
拼写错误的单词等等，这使得
计算机几乎无法识别出答案是什么。这个问题可以通过图形方式进行操作，以便它在视觉上更难以像平常一样弄清楚
验证码。
除了改变一个单词中字母的顺序可以故意地改变字母，以便这个单词没有意义，但在上下文中对于人类来说并不困难。甚至可以使用糟糕的语法和一些其他的东西来使它变得更加困难。
我以为可以创建一个非常大的数据库
网站可以使用的不同语言。
你们怎么想？
Jon
I was thinking about a way to use a similar idea but to dramatically
increase the ability to thawrt bots.
The idea is to use the idea that humans can use context and patterns to
understand much better than computers in such cases as the following,
http://www.simplebits.com/notebook/2...sed_wdors.html
So the idea is to ask the user a very simple questions that is written using
misspelled words and such so that it makes it almost impossible for
computers to recognize what the answer is. The question can be graphically
manipulated so that its also visually harder to figure out like a normal
captcha.
Besides changing the order of the letters in a word one can deliberately
change letters so that the word doesn''t make sense but in the context is not
difficult for a human. One can even use bad grammar and a few other things
to make it more difficult.
I was thinking that one could create a very large database of such things in
different languages that sites could use.
What do you guys think?
Jon
很酷的想法！
不要轻易放弃图片，因为我觉得它会
然后编码人员仍然可以将图像解码为纯文本
（就像他们现在一样）并强制使用每个可能的字母组合
对你的数据库，如果任何可以用Paypal付款的人可以在数据库副本上获得
。或者更好的是，MySQL''LIKE''
与中间的通配符进行比较。
我认为人类验证图像中的另一个重要因素是
网站之间的区别。使用相同的
CAPTCHA代码的所有phpBB论坛都会更容易编写一个脚本，这可以欺骗所有
，而不是验证图像脚本的许多发行版
的行为略有不同，具体取决于它的环境，比如管理网站的人只需管理网站就可以制作自己的背景，或者输入更多
的变化一个秘密的单词或数字，计算
将基于。
祝你好运！
-Mike PII
Cool idea!
Don''t let up on scrambling the image, though, because I think it would
then still be possible for coders to decode the image into plain text
(like they do now) and bruteforce every possible mix of the letters
against your database, if anyone who can pay you with Paypal can get
their hands on a copy of the database. Or better yet, a MySQL ''LIKE''
comparison with a wildcard in the middle.
I think another important factor in human verification images would be
differentiation between websites. All phpBB forums using the same
CAPTCHA code would be easier to write a script for that can fool all
of them than many distributions of a verification image script that
acts slightly different depending on it''s environment, like whoever
manages the website can make their own backgrounds, or add some more
variation by entering a secret word or number on which calculations
will be based.
Good luck!
-Mike PII
" Jon Slaughter" < Jo *********** @ Hotmail.com写信息
新闻：Au *************** @ newssvr19.news .prodigy.net。 ..
"Jon Slaughter" <Jo***********@Hotmail.comwrote in message
news:Au***************@newssvr19.news.prodigy.net. ..
>我正在考虑一种方法来使用类似的想法，但是要大大增加解冻机器人的能力。
这个想法是使用人类可以使用上下文和模式来理解比计算机更好的理解，如下所示：
http://www.simplebits.com/notebook/ 2 ... sed_wdors.html
因此，我们的想法是向用户提出一个非常简单的问题，即使用拼写错误的单词写成
因此，它几乎不可能让b
计算机识别答案是什么。这个问题可以通过图形方式进行操作，以便它在视觉上更难以像平常一样弄清楚
验证码。
除了改变一个单词中字母的顺序可以故意改变字母，以便这个单词没有意义，但在上下文中对于一个人来说并不困难。甚至可以使用糟糕的语法和一些其他的东西来使它变得更加困难。
我以为可以创建一个非常大的数据库用网站可以使用的不同语言的
。
你们怎么想？
>I was thinking about a way to use a similar idea but to dramatically
increase the ability to thawrt bots.
The idea is to use the idea that humans can use context and patterns to
understand much better than computers in such cases as the following,
http://www.simplebits.com/notebook/2...sed_wdors.html
So the idea is to ask the user a very simple questions that is written
using misspelled words and such so that it makes it almost impossible for
computers to recognize what the answer is. The question can be graphically
manipulated so that its also visually harder to figure out like a normal
captcha.
Besides changing the order of the letters in a word one can deliberately
change letters so that the word doesn''t make sense but in the context is
not difficult for a human. One can even use bad grammar and a few other
things to make it more difficult.
I was thinking that one could create a very large database of such things
in different languages that sites could use.
What do you guys think?
你是指责诵读困难的人。
-
Ra ********* @ gmail.com
"好茶。漂亮的房子。 - Worf
You''re giving the finger to dyslexic people.
--
Ra*********@gmail.com
"Good tea. Nice house." -- Worf
" Rami Elomaa" < ra ********* @ gmail.com写信息
新闻：jY ******************** @ reader1.news.saunalaht i.fi ...
"Rami Elomaa" <ra*********@gmail.comwrote in message
news:jY********************@reader1.news.saunalaht i.fi...
" Jon Slaughter" < Jo *********** @ Hotmail.com写信息
新闻：Au *************** @ newssvr19.news .prodigy.net。 ..
"Jon Slaughter" <Jo***********@Hotmail.comwrote in message
news:Au***************@newssvr19.news.prodigy.net. ..
>>我正在考虑一种方法来使用类似的想法，但是要大大增加解冻机器人的能力。
我们的想法是使用人类可以使用上下文和模式来比在下列情况下更好地理解计算机的想法，
http://www.simplebits.com/notebook/2...sed_wdors。 HTML
因此，我们的想法是向用户提出一个非常简单的问题，这些问题是使用拼写错误的单词写成的，这样才能使计算机几乎不可能认识到答案是什么。这个问题可以用图形方式操作，这样它在视觉上也更难以像正常的验证码那样弄清楚。
除了改变单词中的字母顺序之外，还可以故意<改变字母，使这个词没有意义，但在上下文中对人类来说并不困难。甚至可以使用糟糕的语法和一些其他的东西来使它变得更加困难。
我在想，可以用不同的语言创建一个非常大的数据库那些网站可以使用。
你们觉得怎么样？
>>I was thinking about a way to use a similar idea but to dramatically
increase the ability to thawrt bots.
The idea is to use the idea that humans can use context and patterns to
understand much better than computers in such cases as the following,
http://www.simplebits.com/notebook/2...sed_wdors.html
So the idea is to ask the user a very simple questions that is written
using misspelled words and such so that it makes it almost impossible for
computers to recognize what the answer is. The question can be
graphically manipulated so that its also visually harder to figure out
like a normal captcha.
Besides changing the order of the letters in a word one can deliberately
change letters so that the word doesn''t make sense but in the context is
not difficult for a human. One can even use bad grammar and a few other
things to make it more difficult.
I was thinking that one could create a very large database of such things
in different languages that sites could use.
What do you guys think?
你的手指是诵读困难的人。
You''re giving the finger to dyslexic people.
和？所以你认为他们也可以阅读正常的验证码吗？如果他们患有阅读障碍症，那么他们很可能会更容易阅读它们，因为有了b / b $ b $背景...所以你不必阅读信函而只是得到
的问题。
Dyslexia是愚蠢人士的借口。
And? So you think they can read normal captchas too? if they are dyslexic
then chances are they can actually read them easier because there is
context... so you don''t have to read letter for letter but just get the
question.
Dyslexia is an excuse for stupid people.
这篇关于验证码的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
