Suggestions as follows:

There is a Sonicwall Hardware which filters all the content and internet access. It acts as a firewall

Twitter: Twitter API is available only which can Detect and find the nature of tweet
Facebook: As if now Facebook Doesn't Provide its API i.e Without the permission of FB or User you can access his status,pics or videos.

YouTube: YouTube Already have Filtering Which even Detect piracy.

Best Thing we can do is to create a website to be embed in FB,Tweeter and all social media which give an option of span followed by information
This information will be processed by government.

For image filtering we need to use artificial intelligence or we can use a platform like https://www.mturk.com/mturk from Amazon

The data for digital content is mostly unstructured and is difficult to analyse.As we are moving towards the BigData revolution , it should be better to leverage the already existing BigData tools in the market.
One of the most popular tools that I have come across for reading unstructured data is Splunk , some other tools can be found at http://alternativeto.net/software/splunk/.This tools need to be integrated with the website and then the filtering mechanism can be used.

There are numerous tools for content filtering. Every Firewall or Antivirus has parental control filter. All these tools work on a similar concept. They search "Bad words", They track IP address of inappropriate sites and They also track consumer behavior. All this can be done using JavaScript. But we need to think if we can really rely on these tools. Also, Citizens are concerned about net neutrality these days. What we need to do is to create our own system to filter these things.

Rather than using tool, suggesting to develop tool internally by govt with all required security in place of contents, using cloud and big data to process such contents (which can be in various forms images,texts) and filter our predefined set of key words

1) A database consisting of potentially dangerous words needs to be created.
2) As the web traffic passes from the client to the web or application server. In the application server we can implement a black list of the database in step 1. This black list consists of URL encoded,asci and all encoding and machine learning forms. So , when anybody writes thoso wors, the application server blacklists that word or content thereby easily censoring the potentially dangerous sentences.

1. Dansguardian : DansGuardian is an award winning Open Source web content filter which currently runs on Linux, FreeBSD, OpenBSD, NetBSD, Mac OS X, HP-UX, and Solaris
http://dansguardian.org/?page=whatisdg
vendor : Smoothwall Ltd (http://smoothwall.com/)
2. CensorNet :

http://www.censornet.com/en/products/censornetpro/features
3. Save Squid
http://www.safesquid.com/

Hello. I have attached a brief PDF document that provides a brief of some of the well integrated web-based social media and content management/monitoring tools. Now most of these solutions come embedded with analytics, security and content management modules, it would be important to first cross check your immediate requirements. If you are interested in having solutions focused only on regulating website access etc, companies like Net Nanny are the best. Please check attachment. Hope it helps.

By using virtual human engineering we can make more secure