Introduction to the Country Profi les
The country profi les that follow offer a synopsis of the fi ndings and conclusions of
OpenNet Initiative (ONI) research into the factors infl uencing specifi c countries ’ decisions to fi lter or abstain from fi ltering the Internet, as well as the impact, relevance,
and effi cacy of technical fi ltering in a broader context of Internet censorship.
These profi les cover a selection of countries in Asia where the ONI conducted technical testing and analysis from 2009 to 2010. Countries selected for in-depth analysis
are those in which it is believed there is the most to learn about the extent and processes of Internet fi ltering.
The ONI employs a unique ‘ ‘ fusion ’ ’ methodology that combines fi eldwork, technical research, and data mining, fusion, analysis, and visualization. Our aim is to
uncover evidence of Internet content fi ltering in countries under investigation. The
ONI ’ s tests consist of running special software programs within countries under investigation that connect back to databases that contain lists of thousands of URLs, IPs,
and keywords. The lists are broken down into two categories: Global lists include URLs,
IPs, and keywords that are tested in every country and that help us make general
comparisons of accessibility across countries. Global lists also provide a ‘ ‘ snapshot ’ ’ of
accessibility to content typically blocked by fi ltering software programs, and they can
help us understand whether particular software programs are being used in a specifi c
context. Local lists are unique for each country and are usually made up of content
in local languages. These are high-impact URLs, IPs, and keywords, meaning they have
content that is likely to be targeted for fi ltering or that has been reported to have been
so targeted.
Our aim is to run tests on each of the main Internet service providers (ISPs) in a
country over an extended period of time — typically at least two weeks on at least two
occasions. Our accessibility depends very much on our in-country testers, and for
security and other reasons we are not always able to perform comprehensive tests,
meaning in some cases we have only partial results on which to base inferences. Our
specially designed software checks access both within the country and from one or
more control locations simultaneously. Anomalies are analyzed, and determinations
The country profi les that follow offer a synopsis of the fi ndings and conclusions of
OpenNet Initiative (ONI) research into the factors infl uencing specifi c countries ’ decisions to fi lter or abstain from fi ltering the Internet, as well as the impact, relevance,
and effi cacy of technical fi ltering in a broader context of Internet censorship.
These profi les cover a selection of countries in Asia where the ONI conducted technical testing and analysis from 2009 to 2010. Countries selected for in-depth analysis
are those in which it is believed there is the most to learn about the extent and processes of Internet fi ltering.
The ONI employs a unique ‘ ‘ fusion ’ ’ methodology that combines fi eldwork, technical research, and data mining, fusion, analysis, and visualization. Our aim is to
uncover evidence of Internet content fi ltering in countries under investigation. The
ONI ’ s tests consist of running special software programs within countries under investigation that connect back to databases that contain lists of thousands of URLs, IPs,
and keywords. The lists are broken down into two categories: Global lists include URLs,
IPs, and keywords that are tested in every country and that help us make general
comparisons of accessibility across countries. Global lists also provide a ‘ ‘ snapshot ’ ’ of
accessibility to content typically blocked by fi ltering software programs, and they can
help us understand whether particular software programs are being used in a specifi c
context. Local lists are unique for each country and are usually made up of content
in local languages. These are high-impact URLs, IPs, and keywords, meaning they have
content that is likely to be targeted for fi ltering or that has been reported to have been
so targeted.
Our aim is to run tests on each of the main Internet service providers (ISPs) in a
country over an extended period of time — typically at least two weeks on at least two
occasions. Our accessibility depends very much on our in-country testers, and for
security and other reasons we are not always able to perform comprehensive tests,
meaning in some cases we have only partial results on which to base inferences. Our
specially designed software checks access both within the country and from one or
more control locations simultaneously. Anomalies are analyzed, and determinations