Publication: CAPTCHAS : debilidades y fortalezas
Loading...
Identifiers
Publication date
2013-06-24
Defense date
2013-07-10
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
En el presente trabajo se tratará el tema de los CAPTCHAS, es decir, esas pruebas que suelen aparecer cuando queremos registrarnos en una página, en una encuesta, descargarnos algún tipo de fichero, etc. La funcionalidad de estos es diferenciar si se está tratando con un ser humano o con una máquina y, de esta forma, evitar que un robot sea capaz de realizar este tipo de acciones en la web.
Las pruebas de los CAPTCHAS se basan en problemas abiertos de Inteligencia Artificial. Por lo tanto, estas pruebas también estarían relacionadas con la rama de la IA, lo que lleva a investigar y trabajar en su seguridad.
Hoy en día, el tema de los CAPTCHAS como prueba de seguridad, no es 100% fiable, y es que, a pesar de los intentos de crear uno, que verdaderamente pruebe que se está tratando con un ser humano, y no con una máquina, no se ha llegado a encontrar dicho CAPTCHA “perfecto”. Existen numerosas herramientas que son capaces de saltar, en cuestión de minutos, las pruebas consideradas más actuales y desarrolladas.
Cada vez los CAPTCHAS son más “complicados”, es decir, no son fáciles de ver y resolver por un usuario cualquiera, a golpe de vista. Esto se hace con el fin de que también sea más complicado de resolver por una máquina, pero la realidad, es que la mayoría de las veces, es ineficiente. En este estudio se verán los diferentes CAPTCHAS que existen hoy en día, con sus principales fortalezas y debilidades. Se verán desde ejemplos muy sencillos, como introducir una determinada cadena de caracteres, hasta los más complicados, tanto para usuarios como para máquinas, que consisten en resolver una fórmula matemática o un puzle.
Además veremos algunas de las principales herramientas anti-CAPTCHAS, y se crearán y comentarán nuevas ideas de CAPTCHAS, que serían, a día de hoy, muy seguros con respecto a estas herramientas que intentan romperlos.
In this paper, we address the idea of CAPTCHAS, they are tests that usually appear when we check into a page, in a survey, download any file, etc. The functionality of these test, is tell you whether you're dealing with a human or a machine, and then, you can prevent robot can perform those actions on the web. CAPTCHA tests are based on open problems in Artificial Intelligence. So, these tests are also linked to the branch of AI, and this leads to research and work on their safety. Today, CAPTCHAS, like security test, are not 100% reliable, despite attempts to create a CAPTCHA, which really proves that we are dealing with a human, not a machine, it has not been found, it has not been created the "perfect" one. There are many tools that are able to break in minutes the most current and developed test. Today, CAPTCHAS are more and more “complicated”, they are not easy to see and resolve any user, at a glance. This is because, developers want CAPTCHAS are difficult for bots too, but actually, this is often inefficient. In this study, it will be seen different CAPTCHAS that exist today, with its main strengths and weaknesses. It will be seen CAPTCHAS simple, like inserting a particular character string, and CAPTCHAS more complicated, for users and machines, which consist of solving a mathematical formula or a puzzle. Furthermore, it will be seen some anti-CAPTCHAS tools, and it will be create and comment new CAPTCHAS ideas. These ideas are, today, very safe with respect to anti-CAPTCHAS tools views.
In this paper, we address the idea of CAPTCHAS, they are tests that usually appear when we check into a page, in a survey, download any file, etc. The functionality of these test, is tell you whether you're dealing with a human or a machine, and then, you can prevent robot can perform those actions on the web. CAPTCHA tests are based on open problems in Artificial Intelligence. So, these tests are also linked to the branch of AI, and this leads to research and work on their safety. Today, CAPTCHAS, like security test, are not 100% reliable, despite attempts to create a CAPTCHA, which really proves that we are dealing with a human, not a machine, it has not been found, it has not been created the "perfect" one. There are many tools that are able to break in minutes the most current and developed test. Today, CAPTCHAS are more and more “complicated”, they are not easy to see and resolve any user, at a glance. This is because, developers want CAPTCHAS are difficult for bots too, but actually, this is often inefficient. In this study, it will be seen different CAPTCHAS that exist today, with its main strengths and weaknesses. It will be seen CAPTCHAS simple, like inserting a particular character string, and CAPTCHAS more complicated, for users and machines, which consist of solving a mathematical formula or a puzzle. Furthermore, it will be seen some anti-CAPTCHAS tools, and it will be create and comment new CAPTCHAS ideas. These ideas are, today, very safe with respect to anti-CAPTCHAS tools views.
Description
Keywords
Seguridad informática, Captchas, Inteligencia artificial, Proceso de imágenes