I do think there is one way this research could help. If we use this software so that people can learn what "normal" speech on the internet is as opposed to predatory speech. At this point in time I don't think we even know what "normal" is. We need an objective bellcurve to start with in my opinion so that we know stuff such as how obscene something is in relation to the normal and we also need to know how often an individual communicates in an obscene manner. Extremes are going to exist on the bellcurve but without that bellcurve then we don't have any objective way to determine sexually predatory speech and we risk labeling the majority as sexually predators and having many false positives.
An important or critical question is "What is normal speech?" and "how obscene (sexually predatory) is the normal internet user"?
Meaning lets assume if we looked at every text message ever sent by everyone on the internet and 80% of people have said something which could be interpreted as sexually predatory. Then in this case we'd know there are a lot of false positives and it would all depend on how we present that data to the public. My concern was that the media would irrationally sensationalize the data and make it seem like 80% of the internet population are pedophiles and rapists based on some logs and statistics and use this as an excuse to witch hunt and completely destroy the internet.
This is based on my own "gut feeling" not fact, is probably that most people on the internet have sent or do send obscene text messages. I'd also say most people probably don't even remember everything they've said on the internet. So just looking at logs virtually everyone is going to appear like a sex predator, whether they actually are or not. This is why I say don't label the persons being analyzed "sex predators" until we actually know what a sex predator looks like as opposed to a false positive joe sixpack talking dirty via text.
The problem is I don't see how we can objectively determine that difference at this time. It's like our level of understanding of this is so naive and so misinformed that we simply don't have a clue. This is similar to the Kinsey research where the statistics being taken are new but at the same time completely disruptive to the narratives that people believe in. People expect speech on the internet to match speech off the internet for example but most people don't communicate the same on the internet as they do off of the internet.
It's worthwhile research and I'm not against it, I just think it will depend on the interpretation of the results and what we try to do with it. I don't think this information should be used at this point in time in trials, or to try and help law enforcement because I think we need more time to actually figure out what "normal" speech is on the internet as opposed to predatory. If we find out that predatory speech is normal then what?
It seems like you're missing the actual point of this, which is to actually refine that process to weed out a lot more false positives, which would lead to LESS abuse than a system which would return more (incorrect) hits.
And you can, in fact, label something predatory while doing research into it if predatory behavior is exactly what you are looking for. They aren't looking for someone who propositions someone, they aren't looking for someone who just hits on someone,but rather someone approaching another person in a predatory fashion with the intent to manipulate and abuse that person into providing sexual favors, or however they are operationalizing the concept of "predatory." The label may only apply within their research, but it's a perfectly valid thing to do.
I don't believe there is a such thing as a predatory text message. I believe there is predatory behavior but you cannot determine that by text messages alone and I think it's fatuous to believe we can. I suppose they are welcome to waste their time and effort but I can't see any productive outcome from this. I also don't think we will all agree on what is or isn't predatory and this is why there shouldn't be labels of this sort.
And, once more, you are missing the point: the point of this is to tel the people who may come off as predators from the ones who are actually predators. To actually label FEWER people predators who aren't really predatory and MORE people who are predatory.
So if you have a person who doesn't actually know they are texting in a predatory manner and this software determines they are, are the individuals labeled predatory going to be alerted to the fact that some software believes they are predatory? I don't see how it's even going to result in a change of language and grammar if this software isn't installed on computers so that users can determine if their own speech is predatory.
My disagreement is in having some authority judge our speech. We should judge our own speech and learn to censor or adapt our own speech to be less obnoxious or to seem less predatory. But to have someone or something else looking up keywords in our speech and not tell us what its conclusions are or how we are being labeled by it, I don't see how anything productive can come from that. I don't think people who communicate in a predatory manner will learn better communication skills and I don't think you can determine a predator entirely by their communication style anymore than you can determine who all the sociopaths are on Slashdot by the words they use. You can get some clues, you can get some hints, but there will always be enough false positives that you'd need more information than just communication style and you'd also need enough logs of them to know the context (you would need years worth).
I don't see why that part is so hard to get - I mean, again, if you just wanted to return as many hits as possible why not just decide any use of certain terms is predatory and call it a day?
You started to get at it yourself - you think this will be full of false positives, and my entire point is that they're trying to find a way to return fewer false positives than might otherwise be returned by any other method of analysis currently used and abused to flag chat logs and other communications.
My problem is we are relying on something as unreliable as chat logs at all for something so serious. I'm not against analysis of chat logs. I do think you can determine some aspects of psychology from language such as if someone is a narcissist or if someone has aspergers etc. But I do not think the degree of accuracy is high enough to actually be used in court and I don't think logs alone can determine what someones behavior would be. This is why I would never label someone a predator from just logs, I would need a lot more evidence but the problem is they are throwing this label around like it's trivial when it's about the worst thi
For the most part the 60s movements were abysmal failures which resulted in the decline of America starting in the 1970s with the destruction of the nuclear family, the destruction of the single parent income, etc.
Utter bullshit. The antiwar movement stopped the Vietnam War. The ecology movement got the EPA instituted, and if you were alive then and lived near any factory you know how bad the environment was. The civil rights movement was similarly successful.
What destroyed the nuclear family was the fact that STDs were no longer fatal thanks to antibiotics, birth control was cheap and effective unlike before in the world's history, and they legalized abortion. Hell, in the '70s women would walk up to me and ask "wanna fuck?"
The women's movement did indeed allow corporations to screw us normal working class stiffs over, but having to pay for the war, coupled with the Arab Oil Embargo of 1974 caused the inflation that had more to do with women joining the workforce in droves than the women's movement did. Wages stagnated (helped by Nixon's wage/price controls) while prices skyrocketed. You can't blame any of the '60s movements on that. Greedy rich people were the cause.
If you were alive during that period, you clearly weren't paying attention.
The Vietnam war was stopped? Yeah and how many wars have there been since that?
The 60s movement caused the war on drugs and the creation of for profit prisons. Where were the freedom fighters of the 60s when Nixon started the war on drugs and when mandatory minimum sentences were institutionalized?
The EPA? How effective is the EPA? They are minimally effective. The 60s radicals of the left served to legitimize the anti-communists of the right. It allowed for the FBI to bring about COINTELPRO. It is best to call the 60s period a period of civil war in the United States which didn't end until the drug war period of the 80s and 90s. And we still have millions of people in prison on drug offenses.
So if the 60s radicals really accomplished so much why hasn't anything gotten better? That is how we should judge them. They fucked up in the 60s which is why the 70s, 80s and 90s and 2000s are even more fucked up. Having the EPA exist but not be effective is still fucked up. Ending the war in Vietnam and bringing the war to American soil is still fucked up.
sorry, but they're not in the same league as their '60s foregenderneutralpersons
I have to call bullshit here... but I'm not saying that as a defense of the current movement, but rather I'm objecting to your idealization of the 60's. All too many baby boomers seem to have a fuzzy, romanticized version of what happened in the 60's.
There was no shortage of bad actors mixed in with more idealistic folks then, just as is the case today. We have, with varying degrees of success, already sugar coated a lot of 60's history. All of the negative aspects you point out in the current movement have analogous issues in the 60's movement.
Of course, there were a lot of good things that happened as a result of the counterculture movements of the 60's. If we pretend there were no such negative aspects to these movements, and then use this optimistic but false dream of the past to condemn modern movements via a flawed comparison to an idealized version of the 60s that never actually existed... then it seems we have missed the entire point of these counterculture movements.
sorry, but they're not in the same league as their '60s foregenderneutralpersons
I have to call bullshit here... but I'm not saying that as a defense of the current movement, but rather I'm objecting to your idealization of the 60's. All too many baby boomers seem to have a fuzzy, romanticized version of what happened in the 60's.
There was no shortage of bad actors mixed in with more idealistic folks then, just as is the case today. We have, with varying degrees of success, already sugar coated a lot of 60's history. All of the negative aspects you point out in the current movement have analogous issues in the 60's movement.
Of course, there were a lot of good things that happened as a result of the counterculture movements of the 60's. If we pretend there were no such negative aspects to these movements, and then use this optimistic but false dream of the past to condemn modern movements via a flawed comparison to an idealized version of the 60s that never actually existed... then it seems we have missed the entire point of these counterculture movements.
For the most part the 60s movements were abysmal failures which resulted in the decline of America starting in the 1970s with the destruction of the nuclear family, the destruction of the single parent income, etc. Now you have to work harder to get less than you got in the 1960s.
See, the great thing about experiments/projects like this is that they can help us understand if our gut instinct about whether or not something can work or not is actually correct.
As I said, I also am skeptical about how well this could possibly work, but it's definitely think its an interesting problem to look at be ause if it IS possible, despite our mutual skepticism, that's pretty damn neat and can lead to learning more things.
Some of the most interesting bits of knowledge we have came about by asking "pointless" questions and doing "pointless" experiments that people said couldn't work.
Not only is it not likely to work but it's likely to be completely abused. If you look hard enough you can find terrorist phrases in almost any sentence. If you look at it the right way even an innocuous sentence becomes sexually predatory. That is the problem I have, it's debating the meaning of "is", and its not going to find sexual predators. It might find thought criminals and people who talk like predators but the problem is by labeling them "sexual predators" there is bias in the experiment itself.
You don't label something predatory and judge when you're trying to do research and solve a problem. You should simply look at the statistics. This experiment is going to be full of false positives, in fact I'm willing to bet the vast majority of people online who talk like predators aren't sex offenders and aren't predators in real life just like you have people on the internet who are internet tough guys or you have trolls on the internet or racists or anything else. At the end of the day there are always going to be people dumb enough to think if someone said something in a log that they are instantly a racist, or instantly a pedophile, or instantly a rapist. These sorts of people make terrible judges but you can bet they will be the ones judging and volunteering to judge.
It's impossible for software to truly understand the meaning of a sentence. How can software determine which meaning is most correct for a certain context for example? A human can try and determine this but a software would have no way to figure it out.
Swap in political activist, opposition party, occupy movement, flash mob, or hackers, and the project doesn't seem so appealing. The goal sounds like they would like to find an engine to which you could feed in a few examples and have a few thousand computers watching all conversations on the net.
1) Create software the 'statistically identifies' pedos. 2) Swap in patterns for political activist, opposition party, occupy movement, flash mob, or hackers 3) Send in the SWAT team.
Got a problem with that? Sounds like we just identified another pedo with 95% confidence.
edit: captcha: malign
But you can't actually identify anyone via just logs. So unless they want people to change their words all they are trying to govern is language.
This looks like a not-too-well-prepared excercise as there is absolutely no definition of what they mean with "sexual predator," except that a sexual predator tries to gain some sort of a sexually-loaded response from the other side. The problem: what is considered a "sexually-loaded response," would e.g. a boyfriend asking his girlfriend for a bikini-picture qualify as a "predator" even though the act is perfectly common and acceptable, do they deem there is a possibility of a sexually-loaded conversation that still manages to say within the terms of "good behaviour" or are all sexually-loaded comments and conversations inherently "bad behaviour" etc. etc.
I have a feeling the whole point with this is to use the results for "protect the children" - politics in an effort to score brownie-points.
This is more a sort of thought profiling, or sexual thought profiling. It doesn't actually catch predators, it just catches people who communicate like predators. People can learn to communicate in a more civilized manner but lewd or obscene communication is not the same as child molestation, rape, etc. Also it's impossible to determine which words are sexual and which aren't.
How do we know how Jane and Joe are interpreting each other? Are we going to go into their minds and try to judge their thoughts? The whole idea of that is ridiculous. Someone could send you a URL and that could be considered obscene and now you've electronically raped them with a URL? You use a poor choice of words and that is obscene and now you're a predator?
If you identify the object of the advances is a minor, that would make it a predator
So when two high school students are chatting about sex, then what? This is more complicated than, "Is a minor involved."
But do please try and differentiate between an interesting computer science problem, and an actual government putting such an algorithm into use.
Except that it is practically guaranteed to be put to use by law enforcement, assuming it works at all. This is like saying, "Let's not confuse an interesting experiment in identifying which gun fired some bullet with law enforcement activities!"
And the bad part is it's entirely subjective. So the whole thing is stupid. A bunch of logs shouldn't determine who is or isn't a predator.
What if I am just trying to get laid? Seriously, how does one determine from chat text whether a person is a 'sexual predator' vs. someone who is just looking for a casual hookup? Wouldn't the approach be similar if not identical? I smell a FAIL.
1) If you identify the object of the advances is a minor, that would make it a predator. (Possibly an unaware one, if he doesn't know the age of the object. - but either way the conversation needs to be stopped.)
2) If the object of the advances is rejecting the advances, again that would make a predator. Sure, she MIGHT be playing hard to get. But the one making the advances shouldn't assume that. And in any case a really good algorithm might be able to tell the difference between a flirtations "no", and a "FUCK OFF: NO!".
But do please try and differentiate between an interesting computer science problem, and an actual government putting such an algorithm into use.
It's more complicated than that. A predator cannot be someone who simply thinks a certain way but someone who acts it out.
Logs represent thoughts, speech, communication, but you cannot determine what someone would actually do by just looking at a bunch of logs. This wouldn't be any more effective than logging who goes to Teen Porn websites to determine who the predators are.
To expand on this a bit.... in a less creepy way.... I remember being an awkward 20 something geek. Never learned how to really "get girls", more than a bit socially inept. so naturally I started reading up on flirting, and dating.... etc...
After a while I came to a conclusion.... the distance between being friendly and hitting on someone is very very short and often as much in the mind of the individual as anywhere. In fact, most of the things that one is told to do "look her in the eye", "pay attention to what she is saying", "make physical contact from time to time"..... its all standard stuff that many people do in the course of normal conversation.
Even beyond my own issues, I recently witnessed an amusing exchange between a couple of friends of mine who started to talk seperately. One of them took their conversations as potentially expressing intertest, the other was agast at the idea, and couldn't believe the first had thought that.
Is it sometimes or often obvious? Sure, more so in text? maybe. However, I am skeptical that such a system will ever really work, when people have a hard enough time making these calls with full information in person.
It's just not possible to determine who is flirting with who and who is getting sexual favors from who when it's all just text. We know there are signals, body language, words people like to use or whatever but no one knows them all. We know sexuality is sufficiently diverse that different topics might affect people sexually but not others. Basically it's just pointless to try to determine who a predator is via logs unless you just want to find the people who act like predators on the internet.
The problem is just talking like a predator on the internet doesn't mean they are a predator IRL. In the example of perverted justice those individuals had to actually try to meet the underage person to be considered a predator. In these instances you have some evidence of intent or action beyond just talking and thinking about it. But if we are going to say "well this person said these obscene words to this other person... so they are a sex predator" is just ridiculous. Obscenity and lewd conduct is not the same as being a predator.
Anyone can have too much to drink one day and be obscene or lewd or inappropriate, but to be a sex predator someone has to actually be serious about what they are doing. You cannot tell the seriousness of a situation from logs alone. You can only figure out that this adult was having nasty thoughts about some minor.
There is no such thing as e-rape. And a log cannot determine who is or isn't a pedophile.
What determines who is or isn't a pedophile is whether or not the person who generated the log is a sex offender. If they are a sex offender and they are caught talking to minors online it's entirely different because in their case they actually have a history of being a predator.
I think it's pointless and stupid to try to use logs alone to try and determine who will become a predator. I think that is impossible. A log at best can represent a thought, but it doesn't represent an action. What people say and think is not the same as what people do. For instance we could probably find a log of just about anyone to make them appear racist, but that doesn't mean they actually are racist and I don't think we should go analyzing logs to try and see who is and isn't.
Actions are what determine how a person really is.
This is all subjective. If they haven't met in person or tried to meet in person then all you have is subjective interpretation of logs. I don't think logs can make someone a predator. I don't think people take what they say online as if they are saying it under oath or with a degree of seriousness. Sexual favors is also extremely vague of a concept. Anything could constitute a sexual favor so how exactly do we know what sexual favors are appropriate or inappropriate? I just think it's pointless to pour over logs for stuff like this unless you know based on some real evidence that the individual is a previous sex offender.
That's possible when the story is small. But that judge doesn't have enough friends to swing something like this, sport.
The very fact that we're talking about this makes it impossible for that judge to pull anything like that. Do you think Governors are going to bow down to a retired judge? I assure you, if the governor orders it... that kid will be tried for murder. I say tried... not convicted. Getting false convictions in the US is of course also possible but it either requires an incompetent defense, a stupid jury, or a corrupt prosecution. And even then the system is weighted specially in capital crimes to find people not guilty unless the case is especially compelling. Yes, there are terrible miscarriages of justice. But if they were the norm the whole justice system would be a waste of time and we would be better served just lynching people.
While you're accusing me of racism... tell me how much you approve of lynching... it will make the irony a little sweeter.
It depends on the judge. Some judges have more power than governors. It depends on who that judge has dirt on, who owes that judge favors, who is loyal to that judge, the whole patronage and corruption system usually revolves around judges.
So to think this particular judge doesn't have the power to influence a governor is silly when the judge probably was in government far longer and may have far more friends.
Yes but not racial, class. I know enough about the justice system to know when someones dad is a judge they will receive special treatment. Just like if your mother was an exec for IBM you might receive special treatment.
Really? So your contention is that they're letting a Hispanic (democrat - cited for political irony) male go after murdering a black man because they're corrupt? Explain how this benefits the police? Do you think they were bribed to release him? And why would the State's Attorney General not overrule them? Oh he's corrupt too? So what you're then arguing is that there is a big conspiracy with thousands of people in it all trying to let this one Hispanic man off for murdering a black person?
WHY? Why would anyone care enough to go to that much trouble?
Make sense.
His dad is a judge. That is all I needed to know to believe corruption is the problem.
I can kick your door down as a police office and claim I heard someone screaming inside or search your car because I smelled marijuana. You can prove that there was no woman in the house or that there was no marijuana in your car but you can't prove I didn't hear or smell that. And all I need to do is say that and I have probable cause.
Probable cause is one of the many loopholes we give the executive that allows them to do pretty much whatever they want so long as they're willing to follow procedure. So you can try someone for just about anything of you can make a case for it. If those person was shot in cold blood then there's a case. If the police believe he was just defending himself then you can probably still make a case it will just be a waste of time because he will be acquitted.
If the police don't think there is a case that means either there is more to this story then you believe and it's all but certain he was defending himself. Or they're wildly corrupt/incompetent. Yes, Florida does have unusually strong self defense laws. However, murder is not legal either. The contention of the "hoodie" movement is that Mr Martin was murdered. That's illegal and you can form a case around that. If half of what the media believes is true then there is a case. If there isn't case it means the media is more then half wrong... possibly completely wrong.
For example, many black americans want massive wealth redistribution. That is the wealth of millions would be arbitarilly ceased and given to black people merely because they're black. You can say that's an extreme view but with the radicals in the streets these are the views those people are upset about.
It's not that black Americans or anyone is demanding wealth distribution, it's a matter of not enough land for everyone because a few families feel entitled to land they stole generations ago. Land is where wealth comes from, along with hard work. For instance if black people wanted to have farm land what does it take from you?
But it doesn't change the fact that some corporation wants to build a mall, or build something else they feel they are entitled to build, they don't ask the native Americans or blacks or the eco-system itself for permission either. So in a dog eat dog environment where some people are hoarding to the point of destroying the eco-system, while redistribution of wealth isn't the solution neither is hoarding. There has to be enough for everyone and this includes non-human animals, and if you can't understand that then you don't deserve what you have.
There is also the question of profiling. Well, the basis of the profiling complaint is the statistical fact that blacks commit a disproportionate amount of crime
Okay Mr. Racist. Getting caught and arrested a disproportionate amount of times does not prove anything other than you get caught. It proves blacks are more likely to get caught and nothing more. Crimes aren't set in stone and aren't based on anything other than what some lawmakers agree upon to be a crime. If we look at the crimes on the internet for example are most of those criminals going to be black? What about white collar crime? Insider trader? Spying? These are usually considered to be "white" crimes but since no one ever gets caught these crimes don't really count as crime.
. I don't know why that is and can only speculate there are cultural issues. But that is a reality and many black people have jumped to the wrong conclusion that they are merely caught at crime more then white people rather then disproportionately committing the crimes.
It's true. If we make obscenity a crime for instance in an all white neighborhood but we don't make it a crime in an all black neighborhood then suddenly there are more criminals in that white neighborhood than in that black neighborhood. Now let's say all the lawmakers in the white neighborhood who made obscenity a crime were black, and lets say a small elite group of multi-racials decide what is or isn't obscene. What do you predict will happen? The result is more white crime.
This is the problem with idiotic race statistics. You can create criminals by creating new laws which apply to certain cultures more than others to make it look like it's people choosing to be criminals when in reality you just criminalized whatever they were doing. So when file sharing is criminalized, and anal sex, who suffers?
What are we supposed to do with that? Arbitrarily forgive black criminals and release them back into the community? Arbitrarily find random white people guilty of various crimes and lock them up?
If they aren't violent why not? Why arrest people for their culture? If we outlawed hiphop then of course millions of black people would be criminals overnight but I'd be arguing we release all who were arrested as political prisoners just like I'd argue for those who listen to classical or country music to be released. It's wrong to arrest people for their culture or lifestyle. If you don't like the lifestyles of black people or poor people, if you don't want people to be on welfare, sell drugs, or prostitute, then give them some better options or shut up about it, but to arrest them doesn't solve anything.
We're never going to resolve these issues because the co
We know for a fact that he began from a position of safety (since he was on the phone w/ 911)
Being on the phone does not magically make one safe.
We know the victim had no weapon.
Did Zimmernan at the time? In any case, eyewtinesses said they saw Treyvon beating up Zimmerman. Treyvon was a young 6' 3" muscular basketball player. Zimmerman is 5'9", 250lbs, so unlikely to be able to run away. Police confirm Zimmerman had a broken nose. If I was getting beat up by someone who could run me down if I tried to run away, and I had a gun, I would shoot.
We know for a fact that he began from a position of safety (since he was on the phone w/ 911)
Being on the phone does not magically make one safe.
We know the victim had no weapon.
Did Zimmernan at the time? In any case, eyewtinesses said they saw Treyvon beating up Zimmerman. Treyvon was a young 6' 3" muscular basketball player. Zimmerman is 5'9", 250lbs, so unlikely to be able to run away. Police confirm Zimmerman had a broken nose. If I was getting beat up by someone who could run me down if I tried to run away, and I had a gun, I would shoot.
And Witnesses say they saw Zimmerman on top of Trayvon. So which witness do you believe? It's up to a jury to decide and Zimmerman should be arrested so that there can be a trial to answer these questions.
You're right, the general population doesn't control the actions. However, keep two things in mind: the FBI answers to the executive, and the attitude of the executive does color the actions of the FBI. Secondly, the FBI is comprised of parts of the population. To argue that the FBI is somehow an entity that is separate of the public is disingenuous. They vote, they have families, they talk to their neighbors in the private sectors, and live a completely normal life outside of the FBI. The reason that nothing stands in the way of their objectives is that the general public in supports their actions.
You know what the biggest problem was in any terror regime? The fact that it had broad support among the population. Next time there's some horrific abuse of civil liberties going on, ask yourself which one of your positions contributed to it. Chances are, there are at least a few of them.
That is just it they don't live completely normal lives. They live FBI lives in an FBI neighborhood and FBI culture. Normal people don't live anything like how FBI agents live because normal people don't receive security clearances and interact with government agencies. FBI agents have no privacy whatsoever and they know it, while ordinary citizens believe they have privacy and don't know they don't have any. This alone would be a major cultural difference.
As far as civil liberties go, my political positions have nothing to do with what the FBI does. The general public does not support the war on drugs. Look at any poll and you'll see a majority of people are against many of these sorts of decisions. The majority of people do not support some of the reckless behavior of the FBI in fighting the war on terror, this is why the FBI and others try to hide it from the public.
So in the case of the war on drugs the majority of the public just doesn't agree with the objective. And in the case of counter terrorism the majority of the public might be convinced on the objective but does not agree with the draconian methods.
If the emissions leak from the voting machine then anyone can see who you voted for and this could result in people changing their votes. You also have to consider that it might be possible to hack these machines and change votes as well.
There will not be a minimum sentence. Just a minumum for the maximum sentence. Difficult concept, but the idea is that each member state will have a maximum prison sentence of at least 2 years. Judges will be free to sentence someone to a month, if they so choose. Member States can also choose to have a maximum prison sentence of more than 2 years, but not less than 2 years.
Slashdot Mirror
I do think there is one way this research could help. If we use this software so that people can learn what "normal" speech on the internet is as opposed to predatory speech. At this point in time I don't think we even know what "normal" is. We need an objective bellcurve to start with in my opinion so that we know stuff such as how obscene something is in relation to the normal and we also need to know how often an individual communicates in an obscene manner. Extremes are going to exist on the bellcurve but without that bellcurve then we don't have any objective way to determine sexually predatory speech and we risk labeling the majority as sexually predators and having many false positives.
An important or critical question is "What is normal speech?" and "how obscene (sexually predatory) is the normal internet user"?
Meaning lets assume if we looked at every text message ever sent by everyone on the internet and 80% of people have said something which could be interpreted as sexually predatory. Then in this case we'd know there are a lot of false positives and it would all depend on how we present that data to the public. My concern was that the media would irrationally sensationalize the data and make it seem like 80% of the internet population are pedophiles and rapists based on some logs and statistics and use this as an excuse to witch hunt and completely destroy the internet.
This is based on my own "gut feeling" not fact, is probably that most people on the internet have sent or do send obscene text messages. I'd also say most people probably don't even remember everything they've said on the internet. So just looking at logs virtually everyone is going to appear like a sex predator, whether they actually are or not. This is why I say don't label the persons being analyzed "sex predators" until we actually know what a sex predator looks like as opposed to a false positive joe sixpack talking dirty via text.
The problem is I don't see how we can objectively determine that difference at this time. It's like our level of understanding of this is so naive and so misinformed that we simply don't have a clue. This is similar to the Kinsey research where the statistics being taken are new but at the same time completely disruptive to the narratives that people believe in. People expect speech on the internet to match speech off the internet for example but most people don't communicate the same on the internet as they do off of the internet.
It's worthwhile research and I'm not against it, I just think it will depend on the interpretation of the results and what we try to do with it. I don't think this information should be used at this point in time in trials, or to try and help law enforcement because I think we need more time to actually figure out what "normal" speech is on the internet as opposed to predatory. If we find out that predatory speech is normal then what?
It seems like you're missing the actual point of this, which is to actually refine that process to weed out a lot more false positives, which would lead to LESS abuse than a system which would return more (incorrect) hits.
And you can, in fact, label something predatory while doing research into it if predatory behavior is exactly what you are looking for. They aren't looking for someone who propositions someone, they aren't looking for someone who just hits on someone,but rather someone approaching another person in a predatory fashion with the intent to manipulate and abuse that person into providing sexual favors, or however they are operationalizing the concept of "predatory." The label may only apply within their research, but it's a perfectly valid thing to do.
I don't believe there is a such thing as a predatory text message. I believe there is predatory behavior but you cannot determine that by text messages alone and I think it's fatuous to believe we can. I suppose they are welcome to waste their time and effort but I can't see any productive outcome from this. I also don't think we will all agree on what is or isn't predatory and this is why there shouldn't be labels of this sort.
And, once more, you are missing the point: the point of this is to tel the people who may come off as predators from the ones who are actually predators. To actually label FEWER people predators who aren't really predatory and MORE people who are predatory.
So if you have a person who doesn't actually know they are texting in a predatory manner and this software determines they are, are the individuals labeled predatory going to be alerted to the fact that some software believes they are predatory? I don't see how it's even going to result in a change of language and grammar if this software isn't installed on computers so that users can determine if their own speech is predatory.
My disagreement is in having some authority judge our speech. We should judge our own speech and learn to censor or adapt our own speech to be less obnoxious or to seem less predatory. But to have someone or something else looking up keywords in our speech and not tell us what its conclusions are or how we are being labeled by it, I don't see how anything productive can come from that. I don't think people who communicate in a predatory manner will learn better communication skills and I don't think you can determine a predator entirely by their communication style anymore than you can determine who all the sociopaths are on Slashdot by the words they use. You can get some clues, you can get some hints, but there will always be enough false positives that you'd need more information than just communication style and you'd also need enough logs of them to know the context (you would need years worth).
I don't see why that part is so hard to get - I mean, again, if you just wanted to return as many hits as possible why not just decide any use of certain terms is predatory and call it a day?
You started to get at it yourself - you think this will be full of false positives, and my entire point is that they're trying to find a way to return fewer false positives than might otherwise be returned by any other method of analysis currently used and abused to flag chat logs and other communications.
My problem is we are relying on something as unreliable as chat logs at all for something so serious. I'm not against analysis of chat logs. I do think you can determine some aspects of psychology from language such as if someone is a narcissist or if someone has aspergers etc. But I do not think the degree of accuracy is high enough to actually be used in court and I don't think logs alone can determine what someones behavior would be. This is why I would never label someone a predator from just logs, I would need a lot more evidence but the problem is they are throwing this label around like it's trivial when it's about the worst thi
For the most part the 60s movements were abysmal failures which resulted in the decline of America starting in the 1970s with the destruction of the nuclear family, the destruction of the single parent income, etc.
Utter bullshit. The antiwar movement stopped the Vietnam War. The ecology movement got the EPA instituted, and if you were alive then and lived near any factory you know how bad the environment was. The civil rights movement was similarly successful.
What destroyed the nuclear family was the fact that STDs were no longer fatal thanks to antibiotics, birth control was cheap and effective unlike before in the world's history, and they legalized abortion. Hell, in the '70s women would walk up to me and ask "wanna fuck?"
The women's movement did indeed allow corporations to screw us normal working class stiffs over, but having to pay for the war, coupled with the Arab Oil Embargo of 1974 caused the inflation that had more to do with women joining the workforce in droves than the women's movement did. Wages stagnated (helped by Nixon's wage/price controls) while prices skyrocketed. You can't blame any of the '60s movements on that. Greedy rich people were the cause.
If you were alive during that period, you clearly weren't paying attention.
The Vietnam war was stopped? Yeah and how many wars have there been since that?
The 60s movement caused the war on drugs and the creation of for profit prisons. Where were the freedom fighters of the 60s when Nixon started the war on drugs and when mandatory minimum sentences were institutionalized?
The EPA? How effective is the EPA? They are minimally effective. The 60s radicals of the left served to legitimize the anti-communists of the right. It allowed for the FBI to bring about COINTELPRO. It is best to call the 60s period a period of civil war in the United States which didn't end until the drug war period of the 80s and 90s. And we still have millions of people in prison on drug offenses.
So if the 60s radicals really accomplished so much why hasn't anything gotten better? That is how we should judge them. They fucked up in the 60s which is why the 70s, 80s and 90s and 2000s are even more fucked up. Having the EPA exist but not be effective is still fucked up. Ending the war in Vietnam and bringing the war to American soil is still fucked up.
I have to call bullshit here... but I'm not saying that as a defense of the current movement, but rather I'm objecting to your idealization of the 60's. All too many baby boomers seem to have a fuzzy, romanticized version of what happened in the 60's.
There was no shortage of bad actors mixed in with more idealistic folks then, just as is the case today. We have, with varying degrees of success, already sugar coated a lot of 60's history. All of the negative aspects you point out in the current movement have analogous issues in the 60's movement.
Of course, there were a lot of good things that happened as a result of the counterculture movements of the 60's. If we pretend there were no such negative aspects to these movements, and then use this optimistic but false dream of the past to condemn modern movements via a flawed comparison to an idealized version of the 60s that never actually existed... then it seems we have missed the entire point of these counterculture movements.
I have to call bullshit here... but I'm not saying that as a defense of the current movement, but rather I'm objecting to your idealization of the 60's. All too many baby boomers seem to have a fuzzy, romanticized version of what happened in the 60's.
There was no shortage of bad actors mixed in with more idealistic folks then, just as is the case today. We have, with varying degrees of success, already sugar coated a lot of 60's history. All of the negative aspects you point out in the current movement have analogous issues in the 60's movement.
Of course, there were a lot of good things that happened as a result of the counterculture movements of the 60's. If we pretend there were no such negative aspects to these movements, and then use this optimistic but false dream of the past to condemn modern movements via a flawed comparison to an idealized version of the 60s that never actually existed... then it seems we have missed the entire point of these counterculture movements.
For the most part the 60s movements were abysmal failures which resulted in the decline of America starting in the 1970s with the destruction of the nuclear family, the destruction of the single parent income, etc. Now you have to work harder to get less than you got in the 1960s.
See, the great thing about experiments/projects like this is that they can help us understand if our gut instinct about whether or not something can work or not is actually correct.
As I said, I also am skeptical about how well this could possibly work, but it's definitely think its an interesting problem to look at be ause if it IS possible, despite our mutual skepticism, that's pretty damn neat and can lead to learning more things.
Some of the most interesting bits of knowledge we have came about by asking "pointless" questions and doing "pointless" experiments that people said couldn't work.
Not only is it not likely to work but it's likely to be completely abused.
If you look hard enough you can find terrorist phrases in almost any sentence. If you look at it the right way even an innocuous sentence becomes sexually predatory. That is the problem I have, it's debating the meaning of "is", and its not going to find sexual predators. It might find thought criminals and people who talk like predators but the problem is by labeling them "sexual predators" there is bias in the experiment itself.
You don't label something predatory and judge when you're trying to do research and solve a problem. You should simply look at the statistics. This experiment is going to be full of false positives, in fact I'm willing to bet the vast majority of people online who talk like predators aren't sex offenders and aren't predators in real life just like you have people on the internet who are internet tough guys or you have trolls on the internet or racists or anything else. At the end of the day there are always going to be people dumb enough to think if someone said something in a log that they are instantly a racist, or instantly a pedophile, or instantly a rapist. These sorts of people make terrible judges but you can bet they will be the ones judging and volunteering to judge.
It doesn't have to know the correct interpretation.
It just has to flag suspicious activity for the local constabulary to review to find local troublemakers.
(Besides, I think you underestimate the ability of software to discern the meaning of sentences).
It's impossible for software to truly understand the meaning of a sentence. How can software determine which meaning is most correct for a certain context for example? A human can try and determine this but a software would have no way to figure it out.
Swap in political activist, opposition party, occupy movement, flash mob, or hackers, and the project doesn't seem so appealing. The goal sounds like they would like to find an engine to which you could feed in a few examples and have a few thousand computers watching all conversations on the net.
1) Create software the 'statistically identifies' pedos.
2) Swap in patterns for political activist, opposition party, occupy movement, flash mob, or hackers
3) Send in the SWAT team.
Got a problem with that? Sounds like we just identified another pedo with 95% confidence.
edit: captcha: malign
But you can't actually identify anyone via just logs. So unless they want people to change their words all they are trying to govern is language.
The funny thing is it would be a complete waste of money because it wouldn't work.
The software would have no way to know the correct interpretation of a sentence.
This looks like a not-too-well-prepared excercise as there is absolutely no definition of what they mean with "sexual predator," except that a sexual predator tries to gain some sort of a sexually-loaded response from the other side. The problem: what is considered a "sexually-loaded response," would e.g. a boyfriend asking his girlfriend for a bikini-picture qualify as a "predator" even though the act is perfectly common and acceptable, do they deem there is a possibility of a sexually-loaded conversation that still manages to say within the terms of "good behaviour" or are all sexually-loaded comments and conversations inherently "bad behaviour" etc. etc.
I have a feeling the whole point with this is to use the results for "protect the children" - politics in an effort to score brownie-points.
This is more a sort of thought profiling, or sexual thought profiling. It doesn't actually catch predators, it just catches people who communicate like predators. People can learn to communicate in a more civilized manner but lewd or obscene communication is not the same as child molestation, rape, etc. Also it's impossible to determine which words are sexual and which aren't.
How do we know how Jane and Joe are interpreting each other? Are we going to go into their minds and try to judge their thoughts? The whole idea of that is ridiculous. Someone could send you a URL and that could be considered obscene and now you've electronically raped them with a URL? You use a poor choice of words and that is obscene and now you're a predator?
If you identify the object of the advances is a minor, that would make it a predator
So when two high school students are chatting about sex, then what? This is more complicated than, "Is a minor involved."
But do please try and differentiate between an interesting computer science problem, and an actual government putting such an algorithm into use.
Except that it is practically guaranteed to be put to use by law enforcement, assuming it works at all. This is like saying, "Let's not confuse an interesting experiment in identifying which gun fired some bullet with law enforcement activities!"
And the bad part is it's entirely subjective. So the whole thing is stupid. A bunch of logs shouldn't determine who is or isn't a predator.
What if I am just trying to get laid? Seriously, how does one determine from chat text whether a person is a 'sexual predator' vs. someone who is just looking for a casual hookup? Wouldn't the approach be similar if not identical? I smell a FAIL.
1) If you identify the object of the advances is a minor, that would make it a predator. (Possibly an unaware one, if he doesn't know the age of the object. - but either way the conversation needs to be stopped.)
2) If the object of the advances is rejecting the advances, again that would make a predator. Sure, she MIGHT be playing hard to get. But the one making the advances shouldn't assume that. And in any case a really good algorithm might be able to tell the difference between a flirtations "no", and a "FUCK OFF: NO!".
But do please try and differentiate between an interesting computer science problem, and an actual government putting such an algorithm into use.
It's more complicated than that. A predator cannot be someone who simply thinks a certain way but someone who acts it out.
Logs represent thoughts, speech, communication, but you cannot determine what someone would actually do by just looking at a bunch of logs. This wouldn't be any more effective than logging who goes to Teen Porn websites to determine who the predators are.
To expand on this a bit.... in a less creepy way.... I remember being an awkward 20 something geek. Never learned how to really "get girls", more than a bit socially inept. so naturally I started reading up on flirting, and dating.... etc...
After a while I came to a conclusion.... the distance between being friendly and hitting on someone is very very short and often as much in the mind of the individual as anywhere. In fact, most of the things that one is told to do "look her in the eye", "pay attention to what she is saying", "make physical contact from time to time"..... its all standard stuff that many people do in the course of normal conversation.
Even beyond my own issues, I recently witnessed an amusing exchange between a couple of friends of mine who started to talk seperately. One of them took their conversations as potentially expressing intertest, the other was agast at the idea, and couldn't believe the first had thought that.
Is it sometimes or often obvious? Sure, more so in text? maybe. However, I am skeptical that such a system will ever really work, when people have a hard enough time making these calls with full information in person.
It's just not possible to determine who is flirting with who and who is getting sexual favors from who when it's all just text.
We know there are signals, body language, words people like to use or whatever but no one knows them all. We know sexuality is sufficiently diverse that different topics might affect people sexually but not others. Basically it's just pointless to try to determine who a predator is via logs unless you just want to find the people who act like predators on the internet.
The problem is just talking like a predator on the internet doesn't mean they are a predator IRL. In the example of perverted justice those individuals had to actually try to meet the underage person to be considered a predator. In these instances you have some evidence of intent or action beyond just talking and thinking about it. But if we are going to say "well this person said these obscene words to this other person... so they are a sex predator" is just ridiculous. Obscenity and lewd conduct is not the same as being a predator.
Anyone can have too much to drink one day and be obscene or lewd or inappropriate, but to be a sex predator someone has to actually be serious about what they are doing. You cannot tell the seriousness of a situation from logs alone. You can only figure out that this adult was having nasty thoughts about some minor.
There is no such thing as e-rape. And a log cannot determine who is or isn't a pedophile.
What determines who is or isn't a pedophile is whether or not the person who generated the log is a sex offender. If they are a sex offender and they are caught talking to minors online it's entirely different because in their case they actually have a history of being a predator.
I think it's pointless and stupid to try to use logs alone to try and determine who will become a predator. I think that is impossible.
A log at best can represent a thought, but it doesn't represent an action. What people say and think is not the same as what people do. For instance we could probably find a log of just about anyone to make them appear racist, but that doesn't mean they actually are racist and I don't think we should go analyzing logs to try and see who is and isn't.
Actions are what determine how a person really is.
This is all subjective. If they haven't met in person or tried to meet in person then all you have is subjective interpretation of logs. I don't think logs can make someone a predator. I don't think people take what they say online as if they are saying it under oath or with a degree of seriousness. Sexual favors is also extremely vague of a concept. Anything could constitute a sexual favor so how exactly do we know what sexual favors are appropriate or inappropriate? I just think it's pointless to pour over logs for stuff like this unless you know based on some real evidence that the individual is a previous sex offender.
That's possible when the story is small. But that judge doesn't have enough friends to swing something like this, sport.
The very fact that we're talking about this makes it impossible for that judge to pull anything like that. Do you think Governors are going to bow down to a retired judge? I assure you, if the governor orders it... that kid will be tried for murder. I say tried... not convicted. Getting false convictions in the US is of course also possible but it either requires an incompetent defense, a stupid jury, or a corrupt prosecution. And even then the system is weighted specially in capital crimes to find people not guilty unless the case is especially compelling. Yes, there are terrible miscarriages of justice. But if they were the norm the whole justice system would be a waste of time and we would be better served just lynching people.
While you're accusing me of racism... tell me how much you approve of lynching... it will make the irony a little sweeter.
It depends on the judge. Some judges have more power than governors. It depends on who that judge has dirt on, who owes that judge favors, who is loyal to that judge, the whole patronage and corruption system usually revolves around judges.
So to think this particular judge doesn't have the power to influence a governor is silly when the judge probably was in government far longer and may have far more friends.
Your prejudice is noted.
Yes but not racial, class. I know enough about the justice system to know when someones dad is a judge they will receive special treatment. Just like if your mother was an exec for IBM you might receive special treatment.
Really? So your contention is that they're letting a Hispanic (democrat - cited for political irony) male go after murdering a black man because they're corrupt? Explain how this benefits the police? Do you think they were bribed to release him? And why would the State's Attorney General not overrule them? Oh he's corrupt too? So what you're then arguing is that there is a big conspiracy with thousands of people in it all trying to let this one Hispanic man off for murdering a black person?
WHY? Why would anyone care enough to go to that much trouble?
Make sense.
His dad is a judge. That is all I needed to know to believe corruption is the problem.
probable cause is incredibly flexible.
I can kick your door down as a police office and claim I heard someone screaming inside or search your car because I smelled marijuana. You can prove that there was no woman in the house or that there was no marijuana in your car but you can't prove I didn't hear or smell that. And all I need to do is say that and I have probable cause.
Probable cause is one of the many loopholes we give the executive that allows them to do pretty much whatever they want so long as they're willing to follow procedure. So you can try someone for just about anything of you can make a case for it. If those person was shot in cold blood then there's a case. If the police believe he was just defending himself then you can probably still make a case it will just be a waste of time because he will be acquitted.
If the police don't think there is a case that means either there is more to this story then you believe and it's all but certain he was defending himself. Or they're wildly corrupt/incompetent. Yes, Florida does have unusually strong self defense laws. However, murder is not legal either. The contention of the "hoodie" movement is that Mr Martin was murdered. That's illegal and you can form a case around that. If half of what the media believes is true then there is a case. If there isn't case it means the media is more then half wrong... possibly completely wrong.
I'd say Florida police are corrupt.
For example, many black americans want massive wealth redistribution. That is the wealth of millions would be arbitarilly ceased and given to black people merely because they're black. You can say that's an extreme view but with the radicals in the streets these are the views those people are upset about.
It's not that black Americans or anyone is demanding wealth distribution, it's a matter of not enough land for everyone because a few families feel entitled to land they stole generations ago. Land is where wealth comes from, along with hard work. For instance if black people wanted to have farm land what does it take from you?
But it doesn't change the fact that some corporation wants to build a mall, or build something else they feel they are entitled to build, they don't ask the native Americans or blacks or the eco-system itself for permission either. So in a dog eat dog environment where some people are hoarding to the point of destroying the eco-system, while redistribution of wealth isn't the solution neither is hoarding. There has to be enough for everyone and this includes non-human animals, and if you can't understand that then you don't deserve what you have.
There is also the question of profiling. Well, the basis of the profiling complaint is the statistical fact that blacks commit a disproportionate amount of crime
Okay Mr. Racist. Getting caught and arrested a disproportionate amount of times does not prove anything other than you get caught. It proves blacks are more likely to get caught and nothing more. Crimes aren't set in stone and aren't based on anything other than what some lawmakers agree upon to be a crime. If we look at the crimes on the internet for example are most of those criminals going to be black? What about white collar crime? Insider trader? Spying? These are usually considered to be "white" crimes but since no one ever gets caught these crimes don't really count as crime.
. I don't know why that is and can only speculate there are cultural issues. But that is a reality and many black people have jumped to the wrong conclusion that they are merely caught at crime more then white people rather then disproportionately committing the crimes.
It's true. If we make obscenity a crime for instance in an all white neighborhood but we don't make it a crime in an all black neighborhood then suddenly there are more criminals in that white neighborhood than in that black neighborhood. Now let's say all the lawmakers in the white neighborhood who made obscenity a crime were black, and lets say a small elite group of multi-racials decide what is or isn't obscene. What do you predict will happen? The result is more white crime.
This is the problem with idiotic race statistics. You can create criminals by creating new laws which apply to certain cultures more than others to make it look like it's people choosing to be criminals when in reality you just criminalized whatever they were doing. So when file sharing is criminalized, and anal sex, who suffers?
What are we supposed to do with that? Arbitrarily forgive black criminals and release them back into the community? Arbitrarily find random white people guilty of various crimes and lock them up?
If they aren't violent why not? Why arrest people for their culture? If we outlawed hiphop then of course millions of black people would be criminals overnight but I'd be arguing we release all who were arrested as political prisoners just like I'd argue for those who listen to classical or country music to be released. It's wrong to arrest people for their culture or lifestyle. If you don't like the lifestyles of black people or poor people, if you don't want people to be on welfare, sell drugs, or prostitute, then give them some better options or shut up about it, but to arrest them doesn't solve anything.
We're never going to resolve these issues because the co
It would be hard not to have one at this point given everyone's insistence on a trial.
You can try someone for anything. I could try you for child molestation... doesn't mean you're guilty of it or that there is even a case for it.
So bring the trial on. And if he's acquitted... by the rules of double jeopardy he can never be tried for that again.
So go for it. If you think you have a case.
No you can't trial anyone for anything. You need probable cause. If they have no probable cause there will be no trial.
Why do you think there will be a trial?
Lets just suspend our opinions until the trial please. I have no idea what happened here and neither does anyone else.
How do we know there will even be a trial?
We know for a fact that he began from a position of safety (since he was on the phone w/ 911)
Being on the phone does not magically make one safe.
We know the victim had no weapon.
Did Zimmernan at the time? In any case, eyewtinesses said they saw Treyvon beating up Zimmerman. Treyvon was a young 6' 3" muscular basketball player. Zimmerman is 5'9", 250lbs, so unlikely to be able to run away. Police confirm Zimmerman had a broken nose. If I was getting beat up by someone who could run me down if I tried to run away, and I had a gun, I would shoot.
We know for a fact that he began from a position of safety (since he was on the phone w/ 911)
Being on the phone does not magically make one safe.
We know the victim had no weapon.
Did Zimmernan at the time? In any case, eyewtinesses said they saw Treyvon beating up Zimmerman. Treyvon was a young 6' 3" muscular basketball player. Zimmerman is 5'9", 250lbs, so unlikely to be able to run away. Police confirm Zimmerman had a broken nose. If I was getting beat up by someone who could run me down if I tried to run away, and I had a gun, I would shoot.
And Witnesses say they saw Zimmerman on top of Trayvon. So which witness do you believe? It's up to a jury to decide and Zimmerman should be arrested so that there can be a trial to answer these questions.
You're right, the general population doesn't control the actions. However, keep two things in mind: the FBI answers to the executive, and the attitude of the executive does color the actions of the FBI. Secondly, the FBI is comprised of parts of the population. To argue that the FBI is somehow an entity that is separate of the public is disingenuous. They vote, they have families, they talk to their neighbors in the private sectors, and live a completely normal life outside of the FBI. The reason that nothing stands in the way of their objectives is that the general public in supports their actions.
You know what the biggest problem was in any terror regime? The fact that it had broad support among the population. Next time there's some horrific abuse of civil liberties going on, ask yourself which one of your positions contributed to it. Chances are, there are at least a few of them.
That is just it they don't live completely normal lives. They live FBI lives in an FBI neighborhood and FBI culture.
Normal people don't live anything like how FBI agents live because normal people don't receive security clearances and interact with government agencies. FBI agents have no privacy whatsoever and they know it, while ordinary citizens believe they have privacy and don't know they don't have any. This alone would be a major cultural difference.
As far as civil liberties go, my political positions have nothing to do with what the FBI does. The general public does not support the war on drugs. Look at any poll and you'll see a majority of people are against many of these sorts of decisions. The majority of people do not support some of the reckless behavior of the FBI in fighting the war on terror, this is why the FBI and others try to hide it from the public.
So in the case of the war on drugs the majority of the public just doesn't agree with the objective. And in the case of counter terrorism the majority of the public might be convinced on the objective but does not agree with the draconian methods.
If the emissions leak from the voting machine then anyone can see who you voted for and this could result in people changing their votes. You also have to consider that it might be possible to hack these machines and change votes as well.
There will not be a minimum sentence. Just a minumum for the maximum sentence. Difficult concept, but the idea is that each member state will have a maximum prison sentence of at least 2 years. Judges will be free to sentence someone to a month, if they so choose. Member States can also choose to have a maximum prison sentence of more than 2 years, but not less than 2 years.
Why not?