Methodology for Parental Controls Test (July 2010)

In the process of open discussion, we selected 9 complex PC protection products having parental control module (for filtering the sites that are unwelcome for children and teenagers) for participation in the test. In addition, a free Microsoft parental control was selected for the test.

As a result, the following products with parental control function took part in the test:

1. Avira Premium Security Suite 10 (10.0.0.542)
2. BitDefender Internet Security 2010 (13.0.21.347)
3. Dr.Web Security Space 6.0 (6.0.0.03100)
4. F-Secure Internet Security 2010 (10.00 build 246)
5. Kaspersky PURE (9.0.0.192)
6. McAfee Internet Security 2010 (10.5.194)
7. Norton Internet Security 2010 (17.7.0.12)
8. Panda Internet Security 2010 (15.01.00)
9. Trend Micro Internet Security 2010 (17.50.1674.0000)
10. Microsoft Windows Live Family Security *

* Microsoft Windows Live Family Security is free and expands the performance capabilities of parental control integrated into Windows 7. Microsoft recommends using Family Security to protect children online.

All tested parental controls were customized to block only pornographic and erotic content including sex education web-sites (e.g. Wikipedia) and adult humor. Control for all the other categories was disabled. If the program suggested profile creating that was ignored. Profile was forcedly created with pornographic and erotic categories only (if necessary).

Platform

The test was performed under Microsoft Windows 7 x64 operating system with all updates installed for the moment of test launch. Internet Explorer 8 that is a standard browser for this operating system was used.

For parallel parental control testing the operation system source image was virtualized with VMWare Player and cloned for every product tested.

Every tested product database was updated on the same day with the minimum possible mistiming (several hours). Virtual machines were updated in alphabetic order. After that, automatic update was forcedly disabled for all the products.

Microsoft Windows Live Family Security was the only exception as it applies for Microsoft servers online and thus always has updated database. That is why this product was the first to be tested to avoid any special advantages for it (till updates are disabled for the rest of the products). All the other parental controls were tested sequentially in alphabetic order.

Preparation of website list

It is quite obvious that it is impossible to test all the parental controls performance for all the unwelcome websites. That is why the most important task for test preparation was to select representative link arrays that can reflect the parental controls work performance for the whole parent population.

The following categories are considered unwelcome for children: violence, drugs, terrorism, gambling, social networks, etc. But for this test we decided to limit ourselves with testing parental controls for pornographic web-sites blocking.

In most cases, children search for unwelcome content using the search engines. That is why we used Top-20 pornographic and erotic search queries in Russian and English (Google statistics) for the test. Then we selected 20 links selected by those search queries in all the three search systems: Google, Bing and Yandex by the marginal conditions:

• 1-10 – the most relevant  (high click probability, low filter breakthrough risk);
• 41-50 – the least relevant (low click probability, high filter breakthrough risk).

Important! When selecting the links, we additionally used the following rules for search works and links selection:

• If we did not manage to find enough relevant links from the first 50 links in Google (with the option “don’t filter the results”) then another more precise keyword was selected. For example, the search query “school girls” was excluded and “pussy” was changed for “wet pussy”.
• Google ignored the links from the same website under the current link (tab-indented).
• If one and the same link occurred in several search engines it was used several times.
• Dead links were excluded. We had to move from the 20^th position down and from the 41^st position up to change them.
• The links were collected without using any security software on PC.
• The search engines settings included searching by English and Russian URLs. When the search engine gave a Russian URL for an English search query it was accepted and listed.

Thus, we collected 800 adult site links - 400 in Russian and 400 in English. The complete search query list is available for vendors’ representatives only on their demand due to censorship.

Testing procedure

All the links received by abovementioned algorithm were tabulated in Excel using the macros that tried to click the links from the allocated boxes. Depending on the content download results or changing the web-page for a safe one, the following verdict was displayed: Skip or Block.

Some tested products like Avira Premium Security Suite, Dr. Web Secure Space and Kaspersky PURE blocked the link click. The manual click was performed for them and both web-pages with replaced content and link click block were considered as a Block (as they were changed for antivirus web-pages in the browser). If the web-site was blocked by inbuilt Microsoft anti-fishing then they were forcedly unblocked.

If the web-site was unavailable docked for the click moment (it is most topical for Bing and web-sites from 41-50) then Block was assigned for the tested product.

If the site is a list of other site links (SEO-gateway) then it was Blocked only if the catalogue first page links were blocked and they could be more than a dozen in some cases. Otherwise, it was Skipped.

If the link brought to COPAA page (I’m 18/21 or I’m under 18/21) then the following logic worked:

1. If the web-page contained erotic (no matter if it was a photo or an image) then it was Skipped and no maturity click was performed.
2. If the web-page contained no erotic then the maturity click was performed. Then Block or Skip followed depending on the result.
3. If the web-page contained no erotic but redirected to a dead site then it was Blocked in accordance with the previous logic.

Internet today is all invaded with pornographic and erotic web-sites, violence, drugs, terrorism and gambling propaganda, obscenity humor and other content that is quite unwelcomed for children and teenagers’ viewing. All sensible parents would like to protect their kids from such unwelcome content. They can do it installing special filters or so-called “parental control” on their computers.

All leading antivirus vendors advertise parental control function in their computer safety complex products. Besides, free parental control function is built in Windows Vista and Windows 7 operating systems.