<h1>Buenas amigos de **Steemit** hoy les traigo un nuevo Post, tratándose de la IA de Google que indentifica frases tóxicas y engañosas con palabras bonitas en internet</h1> <center>![](https://cdn.steemitimages.com/DQmPEpQ2YXQj3BvFL7RCok4FGg2U2oziHLrzebboBY5yNYq/image.png)</center> <sup>Fuente[Link](https://easysmartech.com/wp-content/uploads/2018/09/Google-usar%C3%A1-la-IA-para-luchar-contra-el-material-dea-buso-y-pornografia-infantil.jpg)</sup> **Google** viene desde hace algún tiempo trabajando en un sistema de **inteligencia artificial** capaz de reconocer y por lo tanto castigar de alguna manera los comentarios o post de **odio y ofensa** que circulan en **internet**. Usando un algoritmo que identifica por medio de palabras clave lo que puede contener lenguaje obscena o violento, **Google** analiza la entrada y la considera negativa usando un sistema de porcentaje. El problema es que no es tan difícil así por ahora engañar a este sistema y hacer que no identifique las groserías y las palabras, por ejemplo en un comentario. En el pasado, el simple hecho de escribir las palabras con algún error de ortografía o insertar un carácter diferente, como un punto en medio de la palabra ayudaba bastante en la tarea de engañar la inteligencia artificial. ![](https://cdn.steemitimages.com/DQmQNoQFQ1cm7W2LDdC177aaznLDQeRLriZKwtWo7w5JciF/image.png) <center>![](https://cdn.steemitimages.com/DQmVd2RJzhU4ESQYSR5ESBZQq41y328KyyN9HzyZPfwGCzM/image.png)</center> <sup>*Fuente[Link](https://www.infomendozaonline.com/wp-content/uploads/2018/05/IA-google.jpg)*</sup> Muchas de ellas ya se ha mejorado, aunque todavía está lejos de la perfección, pero una nueva manera de descubrir el algoritmo de **Google** es insertar palabras **positivas** en medio de los comentarios al azar, como **"amor"**. La presencia de palabras así reduce bastante la nota que se da en porcentaje siendo que cuanto más cerca del **100%** más seguro es de que el comentario sea negativo. Así, de la misma forma que cada palabrita añadida en una frase aumenta su nota, **cada palabra positiva la disminuye**, siendo posible engañar la **inteligencia artificial** que podría dejar pasar comentarios horribles enmascarados con términos bonitos. En cualquier caso, identificar el problema no deja de ser un paso en busca de un sistema perfecto, o tal vez más próximo. Y ya nos indica que sólo el análisis de palabras clave no es suficiente. La idea ahora es exactamente que la **inteligencia artificial** sea aún más inteligente para tener una **internet** cada vez menos tóxica. ![](https://cdn.steemitimages.com/DQmQNoQFQ1cm7W2LDdC177aaznLDQeRLriZKwtWo7w5JciF/image.png) <h1><center>Es todo por hoy, Muchas Gracias!</center></h1> <h1><center><div class="pull-right"> <strong>@informatica</strong></center></h1> https://cdn.steemitimages.com/DQmbff1GrUPR3ZArjg6bZ1UiwRh4yw29kvsexqwQieGYS3M/image.png
post_id | 62,268,532 |
---|---|
author | informatica |
permlink | analiza-frases-toxicas-y-enganosas-ia-google |
category | technology |
json_metadata | {"tags":["technology","spanish","post","success","trending"],"format":"markdown","users":["informatica"],"image":["https:\/\/cdn.steemitimages.com\/DQmPEpQ2YXQj3BvFL7RCok4FGg2U2oziHLrzebboBY5yNYq\/image.png"],"links":["https:\/\/easysmartech.com\/wp-content\/uploads\/2018\/09\/Google-usar%C3%A1-la-IA-para-luchar-contra-el-material-dea-buso-y-pornografia-infantil.jpg","https:\/\/www.infomendozaonline.com\/wp-content\/uploads\/2018\/05\/IA-google.jpg"],"app":"steemit\/0.1"} |
created | 2018-09-12 14:42:03 |
last_update | 2018-09-12 14:42:03 |
depth | 0 |
children | 8 |
net_rshares | 3,608,489,851,905 |
last_payout | 2018-09-19 14:42:03 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 2.917 SBD |
curator_payout_value | 0.882 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 2,896 |
author_reputation | 3,575,471,604,506 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 10,000 |
author_curate_reward | "" |
voter | weight | wgt% | rshares | pct | time |
---|---|---|---|---|---|
hr1 | 0 | 53,638,701,661 | 0.02% | ||
sponge-bob | 0 | 16,360,150,020 | 10% | ||
egonz | 0 | 11,660,236,441 | 50% | ||
jhelbich | 0 | 34,909,615,833 | 75% | ||
cyclope | 0 | 23,234,023,653 | 75% | ||
done | 0 | 1,705,316,575,123 | 35% | ||
danielroolz | 0 | 804,782,443 | 100% | ||
youngogmarqs | 0 | 133,371,736 | 0.02% | ||
ordosjc | 0 | 11,983,253,465 | 67% | ||
abrahan414 | 0 | 893,959,878 | 65% | ||
josevillanueva | 0 | 4,648,690,017 | 100% | ||
newmanjosue | 0 | 7,150,496,467 | 100% | ||
wgonz | 0 | 319,549,954 | 69% | ||
comandoyeya | 0 | 6,850,890,876 | 70% | ||
gracerolon | 0 | 2,089,839,756 | 65% | ||
hosgug | 0 | 12,604,189,864 | 50% | ||
fullcolorpy | 0 | 1,625,677,671 | 60% | ||
recidenterrestre | 0 | 2,374,054,009 | 100% | ||
camiachar | 0 | 3,400,672,074 | 100% | ||
heberwords | 0 | 14,634,866,956 | 100% | ||
delirius | 0 | 542,957,938 | 86% | ||
celim | 0 | 324,581,751 | 70% | ||
acerweb | 0 | 334,826,398 | 72% | ||
pepitagold | 0 | 6,457,577,087 | 55% | ||
atebe | 0 | 346,839,826 | 70% | ||
pastorencina | 0 | 11,844,284,613 | 74% | ||
adeljose | 0 | 608,707,074 | 100% | ||
paty2005 | 0 | 491,389,803 | 100% | ||
youdontsay | 0 | 12,450,283,474 | 11% | ||
pdq | 0 | 1,658,936,051,413 | 35% | ||
medaltrot | 0 | 512,595,834 | 100% | ||
yolofussy | 0 | 520,838,037 | 100% | ||
veronicabracho | 0 | 485,320,760 | 100% |
Lo veo bastante difícil de lograr, hay una palabra por ahí que tira abajo todos los conceptos de porcentajes y cuenta de palabras buenas y malas, se llama "ironía" y es tan difícil de detectar que a la mayoría de las personas les es muy complejo reconocerla.
post_id | 62,269,052 |
---|---|
author | hosgug |
permlink | re-informatica-analiza-frases-toxicas-y-enganosas-ia-google-20180912t144846264z |
category | technology |
json_metadata | {"tags":["technology"],"app":"steemit\/0.1"} |
created | 2018-09-12 14:48:45 |
last_update | 2018-09-12 14:48:45 |
depth | 1 |
children | 1 |
net_rshares | 513,165,556,894 |
last_payout | 2018-09-19 14:48:45 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 0.408 SBD |
curator_payout_value | 0.132 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 258 |
author_reputation | 92,611,872,812,879 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 10,000 |
author_curate_reward | "" |
voter | weight | wgt% | rshares | pct | time |
---|---|---|---|---|---|
bytzz | 0 | 49,323,645,162 | 15% | ||
btu | 0 | 463,011,278,004 | 15% | ||
sergino | 0 | 127,502,611 | 0.25% | ||
informatica | 0 | 481,621,477 | 100% | ||
hdu | 0 | 221,509,640 | 2% |
Realmente amigo, eso si sera muy dificil de detectar, Gracias por Comentar
post_id | 62,373,046 |
---|---|
author | informatica |
permlink | re-hosgug-re-informatica-analiza-frases-toxicas-y-enganosas-ia-google-20180913t184253084z |
category | technology |
json_metadata | {"app":"steemit\/0.1","tags":["technology"]} |
created | 2018-09-13 18:42:57 |
last_update | 2018-09-13 18:42:57 |
depth | 2 |
children | 0 |
net_rshares | 0 |
last_payout | 2018-09-20 18:42:57 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 0.000 SBD |
curator_payout_value | 0.000 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 74 |
author_reputation | 3,575,471,604,506 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 0 |
Hay muchas formas de evadir eso, solo escribir con espacios o con signos no se... buen post
post_id | 62,277,678 |
---|---|
author | danielroolz |
permlink | re-informatica-analiza-frases-toxicas-y-enganosas-ia-google-20180912t164654679z |
category | technology |
json_metadata | {"app":"steemit\/0.1","tags":["technology"]} |
created | 2018-09-12 16:46:57 |
last_update | 2018-09-12 16:46:57 |
depth | 1 |
children | 1 |
net_rshares | 472,476,765 |
last_payout | 2018-09-19 16:46:57 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 0.000 SBD |
curator_payout_value | 0.000 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 91 |
author_reputation | 31,866,424,217 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 10,000 |
author_curate_reward | "" |
voter | weight | wgt% | rshares | pct | time |
---|---|---|---|---|---|
informatica | 0 | 472,476,765 | 100% |
Si amigo y eso es lo que ya estan solucionando, Gracias por Comentar.
post_id | 62,373,118 |
---|---|
author | informatica |
permlink | re-danielroolz-re-informatica-analiza-frases-toxicas-y-enganosas-ia-google-20180913t184356742z |
category | technology |
json_metadata | {"app":"steemit\/0.1","tags":["technology"]} |
created | 2018-09-13 18:44:00 |
last_update | 2018-09-13 18:44:00 |
depth | 2 |
children | 0 |
net_rshares | 0 |
last_payout | 2018-09-20 18:44:00 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 0.000 SBD |
curator_payout_value | 0.000 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 69 |
author_reputation | 3,575,471,604,506 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 0 |
The fact that Goog is even thinking about censoring hate speech is a red flag. It is time to leave Goog to rot in a barrel. * The Christian, hates those who kill children. * The Atheist hates Christians who don't approve of their drinking young blood. Both are hate speech and I know which one the AI is looking for so it never comes up on searches. Free speech, a human right given by God, is not to be limited! But they will try. And I will set my search engine to DuckDuckGo.com or zandex.com - Death to goog for attempting to kill my free speech. Nothing new: God said, "And then shall many be offended, and shall betray one another, and shall hate one another." Matthew 24:10 ## Welcome to another fulfilled prophecy!
post_id | 62,279,443 |
---|---|
author | sponge-bob |
permlink | re-informatica-analiza-frases-toxicas-y-enganosas-ia-google-20180912t171527742z |
category | technology |
json_metadata | {"tags":["technology"],"app":"steemit\/0.1"} |
created | 2018-09-12 17:15:30 |
last_update | 2018-09-12 17:15:30 |
depth | 1 |
children | 3 |
net_rshares | 675,685,737,008 |
last_payout | 2018-09-19 17:15:30 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 0.536 SBD |
curator_payout_value | 0.175 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 735 |
author_reputation | 141,978,381,875,679 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 10,000 |
author_curate_reward | "" |
voter | weight | wgt% | rshares | pct | time |
---|---|---|---|---|---|
bytzz | 0 | 51,150,446,835 | 15% | ||
qubes | 0 | 376,579,773,773 | 100% | ||
btu | 0 | 246,939,348,268 | 8% | ||
hamdan12 | 0 | 543,056,340 | 100% | ||
informatica | 0 | 473,111,792 | 100% |
Las intenciones están muy buenas, ¿en que casos se analizan las palabras o frases? En blogspot, artificios, redes sociales?. Sería bueno si algo así funciona. Buen post.
post_id | 62,289,032 |
---|---|
author | jhelbich |
permlink | re-sponge-bob-re-informatica-analiza-frases-toxicas-y-enganosas-ia-google-20180912t194525822z |
category | technology |
json_metadata | {"tags":["technology"],"app":"steemit\/0.1"} |
created | 2018-09-12 19:45:27 |
last_update | 2018-09-12 19:45:27 |
depth | 2 |
children | 1 |
net_rshares | 295,027,026,421 |
last_payout | 2018-09-19 19:45:27 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 0.234 SBD |
curator_payout_value | 0.074 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 175 |
author_reputation | 186,208,713,666,286 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 10,000 |
author_curate_reward | "" |
voter | weight | wgt% | rshares | pct | time |
---|---|---|---|---|---|
bytzz | 0 | 47,496,843,489 | 15% | ||
btu | 0 | 246,939,348,268 | 8% | ||
sergino | 0 | 127,502,611 | 0.25% | ||
informatica | 0 | 463,332,053 | 100% |
Segun tengo entendido seria todo lo que es filtrado por google, en caso de redes sociales es mi duda,Gracias por Comentar
post_id | 62,373,347 |
---|---|
author | informatica |
permlink | re-jhelbich-re-sponge-bob-re-informatica-analiza-frases-toxicas-y-enganosas-ia-google-20180913t184713902z |
category | technology |
json_metadata | {"app":"steemit\/0.1","tags":["technology"]} |
created | 2018-09-13 18:47:21 |
last_update | 2018-09-13 18:47:21 |
depth | 3 |
children | 0 |
net_rshares | 0 |
last_payout | 2018-09-20 18:47:21 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 0.000 SBD |
curator_payout_value | 0.000 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 121 |
author_reputation | 3,575,471,604,506 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 0 |
it's good point @ sponge-bob, with that we would have limitations to our expressions and feelings
post_id | 62,488,935 |
---|---|
author | informatica |
permlink | re-sponge-bob-re-informatica-analiza-frases-toxicas-y-enganosas-ia-google-20180915t050409225z |
category | technology |
json_metadata | {"app":"steemit\/0.1","tags":["technology"]} |
created | 2018-09-15 05:04:09 |
last_update | 2018-09-15 05:04:09 |
depth | 2 |
children | 0 |
net_rshares | 0 |
last_payout | 2018-09-22 05:04:09 |
cashout_time | 1969-12-31 23:59:59 |
total_payout_value | 0.000 SBD |
curator_payout_value | 0.000 SBD |
pending_payout_value | 0.000 SBD |
promoted | 0.000 SBD |
body_length | 97 |
author_reputation | 3,575,471,604,506 |
root_title | "Analiza frases tóxicas y engañosas - IA Google" |
beneficiaries | [] |
max_accepted_payout | 1,000,000.000 SBD |
percent_steem_dollars | 10,000 |