Heinous content censorship

Metraon

Limp Gawd
Joined
Feb 23, 2011
Messages
307
I had an interesting debate today at work about internet censorship and I decided to dig more into this.

What would be a good technological way to prevent someone from hosting hateful content on a personal web server and then sharing it with other people and stay out of the ''radar'' ? How to prevent someone from connecting with a VPN to a server somewhere ?

What good "technological" hardware or software does hosting companies use to prevent this from happening ? Do they ?

Is this something that could be done on a DNS level ? I want to know more unconventional ways than user complaints and not only on corporate side. ISP or Web hosting providers solutions would be nice !

How does Fortigate, SonicWall, etc etc do collect sites and sites of hateful content ? By hand ? With algorithms ?

Could ISP do something on their end to limit this ?

Thanks
 
Last edited:
Why are you trying to usurp people's natural born right to free speech?
 
Its debatable but we dont want to censor it, we were having discussions about it. I am not a ISP, we are chemical engeniering company. Its was dinner conversation.

We were wondering where was the line, what compagnies like Fortinet uses when they are building a database of heinous content.

What I say is its pretty near impossible to block content, even from the ISP side, but in theory if they firewall everything and block everything at the dns level like openDNS it could block most of the common Joe.

Lets say I am a ISP, I could force each of my customers to use my modem and my router. That router have a firewall from preventing from hosting a website and a webfilter of the heinous content, even block proxies and VPN prehaps.

The problem when you are starting to block other services is you may block legitimate traffic and legimate content, thats usurping people right to free speach.

I dont say its practicaly or economicaly viable. I dont say its doable per say, as people will always try to find a way to circonvent this. But we were wondering what could a ISP and Gorvernements do ?

But lets say that you are the President, and it have been universally accepted that you have to block heinous content what are the technicals ways to do this, or prenvent this at the best extent ?

Thanks
 
Lets say I am a ISP, I could force each of my customers to use my modem and my router. That router have a firewall from preventing from hosting a website and a webfilter of the heinous content, even block proxies and VPN prehaps.

At this moment you are out of business in any democratic country. You can't just filter random words, because you will end up censoring pages about history or biology. You cannot filter VPN, because you will end up losing your customers - for example i work from home and without VPN i couldn't do a thing. Blocking hosting a website ? You mean not allowing any port forwarding ? You are out of business in that moment too.
 
ISPs are NOT there to block anything. It should always be up to the end user/company to filter stuff out.
 
At this moment you are out of business in any democratic country. You can't just filter random words, because you will end up censoring pages about history or biology. You cannot filter VPN, because you will end up losing your customers - for example i work from home and without VPN i couldn't do a thing. Blocking hosting a website ? You mean not allowing any port forwarding ? You are out of business in that moment too.

That is why I said, it would not be economically viable. I agree that a combinaison of words could mean nothing for acutal filtering solutions, and I am not even getting in the language barrier.

IBM Watson is a computer project to uderstand human language by a comptuter and prehaps it could be utilised in the future for this

Thats why I am getting curious as manual blocking would be insanely long. Some of my coworkers says thats pretty much the only "solution".
 
ISPs are NOT there to block anything. It should always be up to the end user/company to filter stuff out.

Yeah, we all pretty much agreed that companys have their responsability and its "simple" for them/us to block content. With GPOS, web filtering, audits, etc etc

The discussions was about more maliciously intended users like heinous groups or pedophiles. What can stop them from hosting a home webserver and from building an "underground" community. The responsabilty of controlling this has to fall to something or someone, some people I work with that liberty of speech cannot be that excusable.


As we have a lawyer in the team, he says, the ISP is a service compagny that should to a certain level control the content on his network
 
Last edited:
technically ISP's can not filter out such things and if they do they could be fined or sued for blocking such things. I know some other countries do this and able to get away from it but here in the states it wont happen due to freedom of speech act.
 
The discussions was about more maliciously intended users like heinous groups or pedophiles. What can stop them from hosting a home webserver and from building an "underground" community. The responsabilty of controlling this has to fall to something or someone, some people I work with that liberty of speech cannot be that excusable.
As long as it remains speech, it's their natural right. There are limitations to this speech, but it falls in the range of personal safety. Can't yell fire in a crowded room, for instance.

Technically, you can do things like deep packet inspection, bayes site inspection and keyword blocking. However, the better question is why we would? There's this belief that society needs to be protected from information. This is incorrect. Society needs to be exposed to as much information as possible to force them to grow up and be able to handle that information.

We can treat adults like children and take away their responsibilities. Adults need to act like adults and learn how to deal with disturbing information in a healthy manner.
 
LEt's start by defining what the "Hateful Content" you are referring to. Is it just wacky web comment posts or blogs posts, etc? or are you including specific websites etc? Websites are commonly categorized for the possibility of filtering. Almost all web-filter applications, integrated into a firewall or as a standalone server, will start out with the categorization system. From there, they allow for fine tuning from there with abilities or blocking individual websites and/or specific words/phrases. It's up the the webfilter administrators and/or policies of the each organization what is filtered and what isn't. Schools in general have very tight requirements for filtration by law, where at the business level filtration may be looked upon more as a productivity tool.

I think the confusion exists between the categorization and the plain language (keyword) blocking.
 
This thread is about how censorship is accomplished, not what is being censored, so let's not get sidetracked by that, as tempting as it may be.

OP, here's an interesting article about how China censors the Internet without completely blocking it. Since they have an essentially unlimited governmental workforce to draw upon, they get to utilize the brute force method of having - at the very least - 100,000 workers performing censorship duties 24/7. That is on top of whatever automated functions they have.
 
This thread is about how censorship is accomplished, not what is being censored, so let's not get sidetracked by that, as tempting as it may be.

But the problem is the methods often depend on the content to begin with. What is considered "hateful content"? A string of words? A specific website? Specific images? What about a word within an image?

By trying to guide the debate, you in effect are attempting to censor me, which is exactly what this thread is about. In this case the method is a redirecting comment rather than an outright block, but the concept is the similar. Given specifics on WHAT needs to be filtered, the HOW's become more apparent. Back to the OP, it sounds like your discussion touched the UK's opt-out national censorship. At that level the "hateful content" will be added to block lists as it's found. And from what the articles I've read state, that the ISPs will be responsible for filtering based on that.
 
Anyway the discussion got no way near civilised after a few days so we left off but thanks to all !
 
By trying to guide the debate, you in effect are attempting to censor me, which is exactly what this thread is about.

Uh, no. If I were attempting to censor you, I would be abusing the report system to try to get an admin to delete your post, or attempting to become an admin so I could delete your posts myself.
 
technically ISP's can not filter out such things and if they do they could be fined or sued for blocking such things. I know some other countries do this and able to get away from it but here in the states it wont happen due to freedom of speech act.
What statute enables the fines, and what are they, specifically? Who administers the fines? Can you provide examples of ISPs that have been fined for filtering content?

The problem when you are starting to block other services is you may block legitimate traffic and legimate content, thats usurping people right to free speach.
Except that "free speech" doesn't extend to the use of the internet. At some point, you're using a service (and maybe equipment, too) that someone else owns. In the agreement that governs that contract, the provider asserts the right to terminate the contract at any time, and the consumer promises not to do anything illegal. When a consumer borrows equipment, they're playing by the owner's rules. Paying for the use of the equipment or service doesn't give the consumer unlimited rights to the service. If they're going over bandwidth limits, abusing servers, or doing other things nefarious, the provider can drop them to defend the service and guarantee quality for other users.

This practice doesn't impinge on anyone's "right to free speech" as the consumer agreed to the terms when they signed up and implied their consent when they continued to use the service. The provider's right to enjoy their equipment trumps the consumer's desires as a user of that service.

Getting someone kicked off a forum has nothing to do with free speech or censorship, either. It's really remarkable that people cry "censorship" or "free speech" when things like that happen when the evicted is able (and welcome!) to continue their behavior elsewhere. If someone in a restaurant is screaming and yelling and asked to leave -- even forcibly removed -- it's not censorship and isn't a violation of free speech. It's the maintenance of order.

I dont say its practicaly or economicaly viable. I dont say its doable per say, as people will always try to find a way to circonvent this. But we were wondering what could a ISP and Gorvernements do ?
Have you ever thought of proofreading your posts, by the way? But, look: it's completely economically viable. People pay for such services, even in democracies. SonicWall, NetNanny, Barracuda Networks and others provide equipment that does filtering. ISPs for those with puritan beliefs make tons of money in a competitive market. Even the much vaulted opendns.com offers filtering for home and business users.

But lets say that you are the President, and it have been universally accepted that you have to block heinous content what are the technicals ways to do this, or prenvent this at the best extent ?

Thanks

Mostly, the services work by blocking traffic to addresses that are deemed unacceptable. Obviously, the definition of "unacceptable" is left to the eye of the beholder; and therefore the application of the filter is arbitrary. When a client machine requests DNS resolution, a list is checked to see if the site is unacceptable. If it is, access is denied. If a client machine requests a connection by IP address, a reverse DNS lookup is performed. If the address is on the unacceptable list, then the request is denied.

The list can be built in many different ways. Manually compiling it from scratch is unrealistic, but since these services monitor large streams of traffic, monitoring the traffic itself to sniff out what might be unacceptable is a pretty common application of machine learning. There are voluntary services that tag pages and sites with questionable content; the RTA Label and the ICRA Label are probably the two most popular. As traffic goes by, the ISP can sniff for those tags. They can also work with the label providers to access their database. Further, they can devise their own standards.

Any successful effort will apply multiple approaches to improve coverage and reduce false alarms.
 
Back
Top