User Guide: OONI Explorer
Last updated: 15th March 2023
OONI Explorer is an open data resource on internet censorship around the world.
This user guide provides step-by-step instructions on how to use OONI Explorer to investigate internet censorship worldwide.
What you will get from this guide:
Learn how to use OONI Explorer to:
Discover blocked websites and apps around the world
Access measurement data that can serve as evidence of internet censorship
Generate charts based on aggregate views of OONI data
Compare censorship across countries and networks
Filter measurements based on various parameters (such as country, ASN, date range, OONI Probe test, websites categories, domains, etc.)
Upon reading this guide, we hope you will feel empowered to investigate internet censorship through the use of OONI Explorer. We also recommend reading the article “Is the Website Blocked? Verifying Internet Censorship with OONI Explorer” written by our partner, Sinar Project.
The following table of contents should help with navigating the guide.
About OONI Explorer
What is OONI Explorer?
OONI Explorer is an open data resource on internet censorship around the world. It hosts more than a billion OONI measurements collected from 241 countries and territories since 2012. Every day, hundreds of thousands of new OONI measurements are openly published on OONI Explorer from around the world in near real-time.
OONI Explorer was created by the Open Observatory of Network Interference (OONI), a non-profit born out of the Tor Project in 2012. OONI Explorer was first launched in March 2016, and has been revamped since.
What data does OONI Explorer host?
Data hosted on OONI Explorer is contributed by OONI Probe users around the world. Therefore, OONI data is OONI Probe test results.
OONI Probe is free software designed to measure internet censorship and other forms of network interference. With the OONI Probe app (available for both mobile and desktop platforms), you can run network measurement tests to check the following:
As soon as you run OONI Probe, your test results are automatically published on OONI Explorer in near real-time.
OONI openly publishes OONI Probe test results gathered from around the world to:
- Increase transparency of internet censorship worldwide;
- Share evidence of internet censorship and other forms of network interference;
- Enable the independent verification of OONI censorship findings;
- Support reproducible research;
- Support research, policy, legal, and advocacy efforts around the world;
- Support public debate on information controls.
To enable public access to OONI Probe test results (more commonly referred to as “measurements”), we made them available on OONI Explorer: a web platform that provides charts based on measurement coverage and a search tool for exploring the measurements.
Why use OONI Explorer?
OONI Explorer is likely the largest global open data resource on internet censorship to date. Thanks to OONI Probe users around the world (who contribute measurements), more than a billion network measurements have been collected from 25,000 networks in 241 countries and territories since 2012. All of these measurements are openly available on OONI Explorer.
You can use OONI Explorer to uncover evidence of internet censorship worldwide. Every day, hundreds of thousands of new measurements are published from around the world. OONI Explorer is an expanding public archive on global internet censorship.
Some questions that you can answer via OONI Explorer include:
- Blocking of websites. Which websites are blocked and by which ISPs? How are websites blocked (i.e. which censorship techniques)? How does the blocking of websites differ across countries, ISPs and over time?
- Blocking of instant messaging apps. Are WhatsApp, Facebook Messenger, Signal, and Telegram blocked? If so, how? How does their blocking differ across networks and over time?
- Blocking of circumvention tools. Do circumvention tools, like Psiphon and Tor, work on tested networks? If not, are they blocked?
You can also use OONI Explorer to independently verify our findings (published through our research reports), and to expand upon our research. Through open data, we aim to support reproducible research.
Who is OONI Explorer for?
OONI Explorer is for anyone interested in investigating internet censorship. The platform was built with the following audiences in mind:
- Human rights defenders. Inform your advocacy efforts based on censorship events and trends identified around the world through OONI data.
- Journalists. Enhance the credibility of your stories by referencing network measurement data as evidence of censorship events. Discover unreported censorship stories through OONI data.
- Lawyers. Examine the legality of censorship events identified through OONI data. Consider using OONI data as evidence in court cases.
- Researchers. Explore your own research questions through OONI network measurement data.
- Circumvention tool projects. Inform the development of your tools and strategies based on information about censorship events around the world.
- The public. Keep your government and your Internet Service Providers (ISPs) to account.
What does OONI Explorer include?
OONI Explorer is a web platform designed to enable the investigation of internet censorship worldwide. To this end, OONI Explorer includes several tools and pages:
Search Tool. Filter OONI measurements based on a variety of factors, such as country, ASN (network), date range, OONI Probe test, website categories, and testing status. Discover automatically confirmed blocked websites, and narrow the measurements based on the questions you want to answer.
Measurement Aggregation Toolkit (MAT). Generate charts based on aggregate views of OONI data and track emergent censorship events around the world.
Circumvention Tool Reachability Dashboard. Check if circumvention tools work around the world based on aggregate views of OONI data.
Country pages. Each OONI Explorer country page includes:
- Graph on monthly measurement coverage per OONI Probe test type;
- Reports published by OONI based on the analysis of OONI data from that country (if available);
- Categories of websites (e.g human rights, news media) confirmed blocked in that country (limited to countries where blocking confirmation is automated based on fingerprints);
- Charts with aggregate measurement views of most recently blocked websites (based on OONI data);
- Charts with aggregate measurement views on the recent testing of instant messaging apps (showing if they’re presenting signs of blocking);
- Charts with aggregate measurement views on the recent testing of circumvention tools (showing if they’re presenting signs of blocking).
OONI Explorer home page. View a graph on monthly OONI measurement coverage worldwide, as well as censorship highlights discovered through OONI data.
Learn how to use the OONI Explorer Search Tool and Measurement Aggregation Toolkit (MAT) through the following sections of this guide.
Using OONI Explorer
With the OONI Explorer Search Tool, you can filter OONI measurements based on a variety of factors (such as country, ASN, date range, OONI Probe test, website categories, and testing status), and narrow the measurements based on the questions you want to answer.
Notably, you can easily discover automatically confirmed blocked websites around the world.
You can use the OONI Explorer Search Tool through the following steps:
Step 1. Access the OONI Explorer Search Tool page: https://explorer.ooni.org/search
As you will see, the Search Tool page includes a variety of filters on the left, and a list of OONI Probe test results (“measurements”) from around the world on the right. The most recent OONI measurements are listed first, with each row including an individual measurement.
Each row includes certain metadata that enables you to gain an overview of the measurement. Such metadata includes:
- Country code (e.g
BR for Brazil) and flag (enabling identification of the country from which the measurement was collected); - ASN (enabling identification of the network from which the measurement was collected);
- Date and time of testing (in UTC);
- OONI Probe test name (e.g
Psiphon for the OONI Probe Psiphon test); - Tested URL;
- Testing status (e.g
OK).
For example, through the following measurement (the first listed in the above screenshot), we can see that Psiphon was found reachable (“OK”) when tested in Brazil (AS27699 network) on 7th March 2023 (at 17:30 UTC).
You can learn more (and access the relevant data) by clicking on the measurement row. This is discussed further in the “Measurement pages” sections below.
Filters
As the OONI Explorer Search page lists all OONI measurements from around the world, we provide filters that enable you to narrow your search based on specific parameters.
Specifically, the OONI Explorer Search page includes filters for narrowing the measurements based on:
- Country
- ASN
- Date range
- OONI Probe test
- Website categories
- Domain
- Input
- Testing status
Country
When you first arrive on the Search page, the default Country filter is set to Any, which is why you view measurements from around the world.
To limit OONI measurements to a specific country:
Step 1. Click on the Country dropdown menu and click on a country (in this example, we select Azerbaijan).
Step 2. Click Filter Results at the end of the filters to narrow your search to measurements from the selected country.
You will now only see measurements collected from the selected country (for example, Azerbaijan).
ASN
Internet censorship often differs from network to network within a country. It can therefore be interesting to compare censorship across networks, or to limit your analysis to measurements from a specific network.
By default, the Search page lists measurements from all tested ASNs.
An Autonomous System Number (ASN) is a unique identifier of an autonomous system (AS). This number allows the respective autonomous system to exchange routing information with other systems. An Internet Service Provider (ISP) usually has an officially registered ASN (and they can have more than one ASN). OONI Probe collects the ASN to identify the network in which each test was performed.
Looking up an ASN number in a web search engine (such as Google) will show you which ISP it corresponds to. For example, searching for “AS30722” should return “Vodafone Italia”.
To narrow your search to measurements collected from a specific network:
Step 1. Look up the ASN for the network you’re interested in (for example, MTN Uganda) in a web search engine (such as Google).
The first search results should display the ASN for the provider you’re interested in. In the above example, we can see that by searching “MTN Uganda ASN” we get the following ASN: AS20294.
Step 2. Type the ASN (e.g AS20294) in the ASN field of the Search page filters.
Important: Please ensure that you either have Any or the country where this provider operates (e.g Uganda) selected in the Country filter. If you have selected a country which is unrelated to the added ASN, you will get a “No Results Found” message.
Step 3. Click Filter Results at the end of the filters to narrow your search to measurements from the selected network.
You will now only see measurements collected from the selected network (for example, MTN Uganda).
Date range
OONI data spans for 2012 to date, with new measurements published from around the world in near real-time.
When you first arrive on the Search page, you may notice that only measurements collected over the last month are listed. This default monthly date range has been set for performance reasons.
To change the date range:
Step 1. Click under the From filter to select the starting date. You can select the day and year of your choice, or you can click on one of the Today, Last Week, Last Month, or Last Year buttons (if you would like to narrow your measurement search to the last day, week, month, or year).
Step 2. Click Apply to add the selected starting date. In this example, we have selected 1st February 2022 as the starting date for our measurements.
Step 3. Click under the Until filter to select the end date. In this example, we have selected 31st March 2022 as the end date for our measurements.
Step 4. Click Apply to add the selected end date.
Step 5. Click Filter Results at the end of the filters to narrow your search to measurements for the selected date range.
You will now only see measurements collected from the selected date range (for example, 1st February 2022 to 31st March 2022).
You can further narrow your measurements to a specific country and/or ASN by adjusting the relevant filters (as discussed previously).
Note:
- OONI data is generally available from 2012 to date. However, the availability of data depends on when OONI Probe users in each country started running tests. In many countries, for example, OONI measurement coverage only began in 2016.
- When selecting large date ranges (for example, larger than a year), OONI Explorer might take a bit longer to load the measurements. We thank you for your patience.
- To view measurements from a specific day, please select that day and the next day in the date range (for example, to view measurements from 8th March 2023, select
2023-03-08 in the From filter, and 2023-03-09 in the Until filter).
OONI Probe test
Depending on the question you want to answer, you are probably interested in measurements collected from different OONI Probe tests.
For example, if you’re interested in checking the blocking of WhatsApp, you’d need to check measurements from the OONI Probe WhatsApp test. If you’re interested in checking the blocking of websites, you’d need to check measurements from the OONI Probe Web Connectivity test. You can learn about all the available OONI Probe tests (and how they work) here.
To narrow the measurements to specific tests:
Step 1. Click the Test Name dropdown menu in the Search page.
As you will see through the dropdown menu, OONI Probe tests are categorized based on the type of testing that they perform. For example, the tests for the WhatsApp, Facebook Messenger, Signal, and Telegram apps are listed under the “Instant Messaging” category.
Step 2. Click on an OONI Probe test name. In this example, we clicked on “Web Connectivity Test” to get measurements pertaining to the testing of websites.
Step 3. Click Filter Results at the end of the filters to narrow your search to measurements from the selected OONI Probe test.
You will now only see measurements collected from the selected OONI Probe test (Web Connectivity).
You can further narrow your measurements to a specific country, ASN and/or date range by adjusting the relevant filters (as discussed previously).
Website categories
The OONI Probe Web Connectivity test measures a wide range of websites included in the public, community-curated test lists hosted by the Citizen Lab. Each of the URLs included in these lists is categorized based on 30 standardized category codes.
You can use the OONI Explorer Search Tool to check the blocking of specific categories of websites (such as news media, LGTQI and human rights content) around the world.
To narrow your measurement search to specific website categories:
Step 1. Click the Website Categories dropdown menu in the Search page.
Note: The “Website Categories” filter will only appear if you have “Web Connectivity Test” (or “Any”) selected in the Test Name filter.
Step 2. Click on a website category. In this example, we clicked on “News Media” to get measurements pertaining to the testing of news media websites.
Step 3. Click Filter Results at the end of the filters to narrow your search to measurements for the selected website category.
You will now only see measurements collected from the selected website category (News Media).
You can further narrow your measurements to a specific country, ASN and/or date range by adjusting the relevant filters (as discussed previously).
Domain
You may be interested in checking the availability of a particular website globally, or in specific countries. With the OONI Explorer Search Tool, you can easily check the blocking of particular domains or IP addresses.
To narrow your search to measurements for a specific domain:
Step 1. Type a domain name in the Domain field. For example, twitter.com is the domain of the Twitter website.
Important: Please ensure that the domain name is typed correctly. If it is mis-typed, OONI Explorer will not provide relevant measurements. If you are unsure, please copy-paste the domain when accessing the site from a browser (while excluding the HTTP(S) prefix).
Note: The “Domain” filter will only appear if you have “Web Connectivity Test” (or “Any”) selected in the Test Name filter.
Step 2. Click Filter Results at the end of the filters to narrow your search to measurements for the selected domain.
You will now only see measurements collected from the testing of the selected domain (twitter.com).
You can further narrow your measurements to a specific country, ASN and/or date range by adjusting the relevant filters (as discussed previously).
Note: If you don’t get any measurements for the selected domain, that may be due to one of the following reasons:
With the OONI Explorer Search Tool, you can easily check the blocking of particular URLs or IP addresses.
To narrow your search to measurements for a specific URL:
Step 1. Type a URL in the Input field. For example, https://twitter.com/ is the URL of the Twitter website.
Important: Please ensure that the URL is typed correctly, and that it includes a trailing slash (/). If it is mis-typed, OONI Explorer will not provide relevant measurements. If you are unsure, please copy-paste the URL when accessing the site from a browser (including the HTTP(S) prefix).
Note: The “Input” filter will only appear if you have “Web Connectivity Test” (or “Any”) selected in the Test Name filter.
Step 2. Click Filter Results at the end of the filters to narrow your search to measurements for the selected URL.
You will now only see measurements collected from the testing of the selected URL (https://twitter.com/).
You can further narrow your measurements to a specific country, ASN and/or date range by adjusting the relevant filters (as discussed previously).
Note: If you don’t get any measurements for the selected URL, that may be due to one of the following reasons:
Testing status
Generally, OONI Probe tests will present one of the following three types of results:
- Normal. A “normal” measurement is a testing result with the expected outcome, where nothing is out of the ordinary. When websites and apps are tested, a “normal” measurement is when they are considered accessible. In the case of middlebox tests, a “normal” measurement is when no middleboxes have been detected, and there are therefore no clear signs of network tampering.
- Confirmed blocked. A measurement is “confirmed blocked” when the blocking of a website is automatically detected based on fingerprints. Based on our current heuristics, this can only apply to websites when an Internet Service Provider (ISP) serves a block page (which notifies the user that the website is intentionally blocked) or when DNS resolution returns an IP associated with censorship.
- Anomalous. An “anomalous” measurement is a testing result which is flagged because it presents signs of potential network interference (such as the blocking of a website or app, or the presence of a middlebox). An anomalous measurement does not necessarily contain evidence of internet censorship, as it might be a false positive. Essentially, an anomalous measurement signals that something is wrong and that we should look deeper into the measurement data to determine what is happening.
By default, the OONI Explorer Search Tool presents all measurements, regardless of the testing result.
Note: The “Status” filter will only present the “Confirmed” option if you have “Web Connectivity Test” (or “Any”) selected in the Test Name filter.
To narrow your search to automatically confirmed blocked websites:
Step 1. Click Confirmed in the Status filter of the Search page.
Step 2. Click Filter Results to narrow your search to measurements for automatically confirmed blocked websites.
You will now see measurements from confirmed blocked websites around the world.
You can further narrow your measurements to a specific country, ASN, domain and/or date range by adjusting the relevant filters (as discussed previously).
To narrow your search to anomalous measurements:
Step 1. Click Anomalies in the Status filter of the Search page.
Note: By default, failed measurements (when the experiment failed to perform as expected) are disabled from the view, but you can enable them by unticking the “Hide failed measurements” option. As some failed measurements can be symptomatic of censorship, we encourage researchers to analyze them.
Step 2. Click Filter Results to narrow your search to anomalous measurements.
You will now only see measurements that present anomalies (signs of network interference).
Anomalous measurements pertaining to the testing of websites also list the specific type of anomaly (DNS, TCP/IP, HTTP-failure, HTTP-diff), which can help with characterizing blocking (for example, a DNS anomaly can be indicative of DNS tampering). However, it’s important to look at anomalies in aggregate, and to inspect the raw measurement data to rule out false positives.
Measurement pages
Testing details and evidence of potential internet censorship are available in the measurement pages of each OONI Probe test result.
To access a measurement page:
Step 1. Access the OONI Explorer Search Tool page: https://explorer.ooni.org/search
Step 2. Click on one of the listed measurement rows. In this example, we clicked on the latest WhatsApp measurement collected from Russia.
Each measurement page starts off with a color-coded banner (green for OK, orange for Anomaly, red for Confirmed blocked websites), providing an overview of key information pertaining to the measurement.
In the above example, we can see that WhatsApp was found accessible when tested in Russia on LTD “Erline” (AS47895) on 8th March 2023 (at 6:22 PM UTC).
The banner also includes 3 links:
- Country. By clicking on the country name (e.g Russia), you will access the country-specific page on OONI Explorer, providing measurement information specific to that country.
- ASN. By clicking on the listed ASN (e.g AS47895, LTD “Erline”), you will access a network-centric page, providing measurement information specific to that network.
- Verify. This feature enables (experienced) OONI Explorer users to submit feedback on measurements, enabling us to improve our analysis heuristics.
Step 3. Scroll down the measurement page to view further details.
After the banner, we provide a summary of the testing details (which differ from test to test, depending on how each OONI Probe test works). In the above example (taken from a measurement testing WhatsApp in Russia), we can see that all connections to tested WhatsApp endpoints were successful. We can also see that, overall, the testing of WhatsApp Mobile, WhatsApp Web, and WhatsApp’s registration service were successful. For these reasons, the test determined that “WhatsApp was accessible” at the moment (and on the network) when it was tested in Russia.
To view the measurement details, scroll down to the end of the measurement page, where you can also download the raw data in JSON format.
Step 4. Click Expand All to view the measurement details of the raw data.
Most of the details that are relevant to the measurement results are nested under test_keys.
Step 5. Expand the keys to view further measurement details.
Confirmed
Websites are annotated as “confirmed blocked” when the block is implemented in such a way that we are able to automatically detect it based on fingerprints.
This can occur in two cases:
- A block page is served;
- DNS resolution returns an IP associated with censorship.
Automatically detecting the above cases requires that relevant fingerprints have been added to the OONI database. ISPs also block websites using a variety of different censorship techniques (such as IP blocking, resetting connections, etc), but since those cases aren’t easy to fingerprint (or automatically distinguish from false positives), they are annotated as “anomalies” (discussed in the next section).
By filtering measurements based on “confirmed” in the OONI Explorer Search Tool (explained in previous sections), you can access measurements annotated as “confirmed blocked” from around the world.
Step 1. Click on a row to access a “confirmed blocked” measurement.
In this example, we clicked on a measurement pertaining to the testing of https://twitter.com in Iran on 9th March 2023.
Upon accessing the measurement, the banner informs us that https://twitter.com was confirmed blocked when tested in Iran (on AS50810) on 9th March 2023 (at 10:33 AM UTC).
Step 2. Scroll down the measurement page to view further details.
In the summary of the testing details (illustrated above), we can see that DNS resolution returned the private IPv4 address 10.10.34.35 (instead of Twitter’s actual IP address), which is commonly used by Iranian ISPs for the implementation of censorship. In other words, OONI data shows that users of this network in Iran cannot access Twitter because instead of getting Twitter’s IP address, they get the private IP 10.10.34.35. OONI Explorer displays this case as “confirmed blocked” because the IP 10.10.34.35 is included in the fingerprints added to OONI’s database.
To view the measurement details, scroll down to the end of the measurement page, where you can also download the raw data in JSON format.
Step 3. Click on test_keys to expand the nested measurement details.
This data can potentially serve as evidence of internet censorship, since it shows that the ISP implemented DNS based tampering, returning a private IP (known to be associated with censorship) instead of the actual IP address for the requested website (https://twitter.com).
In other “confirmed blocked” measurements, you will find cases where:
Anomaly
OONI measurements are annotated as “anomalies” when they present signs of potential network interference (such as the blocking of a website or app).
Unlike “confirmed blocked” measurements (which only apply to websites), measurements from most OONI Probe tests can result in anomalies if the tested service fails to meet the testing criteria for accessibility (or if the tested network presents signs of network tampering). Measurements from the testing of instant messaging apps (such as WhatsApp) or circumvention tools (such as Tor) present anomalies if they are blocked. However, false positives can emerge, which is why it’s important to look at anomalous measurements in aggregate (discussed in the next section).
Many cases of website blocking are not automatically detected, and will be present in “anomalous” measurements. Specifically, OONI’s Web Connectivity test (which measures websites for blocking) identifies the following types of anomalies (while automatically comparing measurements from the local, tested network with those from a control vantage point):
- DNS anomaly. If the DNS responses (such as the IP addresses mapped to host names) do not match;
- TCP/IP anomaly. If a TCP session to connect to websites was not established over the network of the user;
- HTTP anomaly.
These anomalies can indicate the presence of DNS tampering (DNS anomaly), IP blocking (TCP/IP anomaly), HTTP blocking (e.g block page), or TLS based interference (e.g connection reset observed right after the ClientHello message during the TLS handshake). However, false positives can emerge due to many reasons. It is therefore important to examine anomalies in aggregate, further aggregate based on anomaly types (e.g HTTP-failure), and to check if the tested service consistently presents the same failure (e.g connection_reset) on the same tested ASN. If a tested service consistently presents the same failures, those measurements provide a stronger signal of blocking.
By filtering measurements based on “anomaly” in the OONI Explorer Search Tool (explained in previous sections), you can access measurements annotated as “anomalies” from around the world.
Step 1. Click on a row to access an “anomalous” measurement.
In this example, we clicked on a measurement pertaining to the testing of https://twitter.com in Russia on 11th March 2023. From the measurement listing, we can see that the testing resulted in an http-failure (indicating that the HTTP experiment failed).
Upon accessing the measurement, the banner informs us that the testing of https://twitter.com presented an “anomaly” (as the HTTP request failed) when tested in Russia (on AS42437) on 11th March 2023 (at 12:51 PM UTC).
Step 2. Scroll down the measurement page to view further details.
In the summary of the testing details (illustrated above), we can see that the DNS queries returned the correct IP addresses, and that the TCP connections to the resolved IPs were successful.
However, the HTTP experiment failed, resulting in a connection_reset error. Specifically, the HTTP request returned no data, because the connection was reset. As a result, a user on this network in Russia likely could not access https://twitter.com.
To view the measurement details, scroll down to the end of the measurement page, where you can also download the raw data in JSON format.
Step 3. Click on test_keys to expand the nested measurement details.
From the test_keys (illustrated above), we can see that the HTTP experiment failed, resulting in a connection_reset error. To explore further:
Step 4. Click on tls_handshakes.
The tls_handshakes key has details nested below. In the above example, there are 4 nested items (one for each of the 4 resolved IPs).
Step 5. Click on the nested items under the tls_handshakes key.
We can see that the connection was reset (connection_reset error) for each of Twitter’s resolved IPs during the TLS handshake, suggesting TLS level interference. To explore further:
Step 6. Click on network_events (and on the items nested under this key).
Through the network_events keys, we can see the TLS handshake process for each of the resolved IP addresses.
In the above example, we can see that the probe successfully connected to the IP 104.244.42.1:443 and started the TLS handshake. However, the connection was reset right after the ClientHello message (read operation) during the TLS handshake. This shows TLS level interference, and may indicate the use of Deep Packet Inspection (DPI) technology. Based on this measurement, access to Twitter appears to be blocked in Russia, though it’s important to look at relevant measurements in aggregate (discussed in the next section).
While individual measurements contain raw data with details that are necessary for characterizing (and confirming) censorship events, it’s important to look at many relevant measurements at once (in aggregate) to identify trends and patterns (and rule out false positives). For example, if the testing of WhatsApp presents anomalies every time it is tested on a network in February 2023, that provides a stronger signal of blocking than if WhatsApp only presented one anomaly, but was found accessible every other time it was tested on that network (i.e that one anomaly may have been a false positive).
To enable you to visualize OONI data and view it in aggregate, we built the Measurement Aggregation Toolkit (MAT). You can use the MAT to generate charts based on aggregate views of real-time OONI data collected from around the world.
Specifically, you can use the MAT to answer the following types of questions:
- Does the testing of a service (e.g. Facebook) present signs of blocking every time that it is tested in a country? This can be helpful for ruling out false positives.
- What types of websites (e.g. human rights websites) are blocked in each country?
- In which countries is a specific website (e.g. bbc.com) blocked?
- How does the blocking of different apps (e.g. WhatsApp or Telegram) vary across countries?
- How does the blocking of a service vary across countries and ASNs?
- How does the blocking of a service change over time?
You can answer such questions and generate charts based on OONI data by adjusting the MAT filters.
Filters
At the start of the MAT page, you can adjust the filters and click Show Chart to plot charts based on aggregate views of OONI data.
The MAT includes the following filters:
- Country: Select a country through the drop-down menu (the “All Countries” option will show global coverage).
- Test Name: Select an OONI Probe test based on which you would like to get measurements (for example, select
Web Connectivity to view the testing of websites). - Domain: Type the domain for the website you would like to get measurements (e.g.
twitter.com). - Input: Type the URL for the website you would like to get measurements (e.g
https://twitter.com/). - Website Categories: Select the website category for which you would like to get measurements (e.g.
News Media for news media websites). - ASN: Type the ASN of the network for which you would like to get measurements (e.g.
AS30722 for Vodafone Italia). - Date range: Select the date range of the measurements by adjusting the
From and Until filters (OONI data spans from 2012 to date, but large date ranges may take longer to load). - Time Granularity: Adjust the time granularity of the measurements (daily, weekly, or monthly for longer time intervals).
- Columns: Select the values (
measurement day, website categories, countries) that you would like to appear on the horizontal axis of your chart. - Rows: Select the values (
domain,website categories, countries, asn) that you would like to appear on the vertical axis of your chart.
Measurements in MAT charts (similarly to the OONI Explorer Search Tool) include the following values:
- OK count: Successful measurements (i.e. NO sign of internet censorship).
- Confirmed count: Measurements from automatically confirmed blocked websites (e.g. a block page was served).
- Anomaly count: Measurements that provide signs of potential blocking (however, false positives can occur).
- Failure count: Measurements from failed experiments (however, they can sometimes be symptomatic of censorship).
Filters for the y-axis
If you select a value in the Row (y-axis) filter (and click Show Chart), a table with additional filters will appear before your charts. This table includes the aforementioned measurement values, along with the overall measurement count.
In the above example, we have selected Website Categories in the Row filter, thus checking the testing of websites based on categories globally.
For each category (during the selected date range), the table lists the:
- Measurement Count: Total number of collected measurements;
- Anomaly Count: Number of measurements that presented anomalies;
- Confirmed Count: Number of measurements where the blocking of websites is automatically confirmed;
- Failed Count: Number of failed measurements.
Important: Only website measurements can be annotated as confirmed. For all other tests, please refer to anomalous measurements, while comparing them with the measurement count (i.e a small percentage of anomalies, in comparison to the measurement count, may indicate false positives).
By clicking on these values in the table, the measurements with the largest volume will be listed first. For example, by clicking Confirmed Count, the category News Media is listed first with the largest volume of confirmed blocked measurements out of all categories (during the selected date range globally).
Note: The above example does not necessarily mean that news media websites are actually blocked more than other website categories globally. This finding is limited by several factors, such as the number and types of websites tested (and the related selection bias) and measurement coverage.
To generate MAT charts, select the categories you would like to display (e.g News Media and Human Rights Issues) and click Apply.
Generating charts based on OONI data
To generate MAT charts based on aggregate views of OONI data:
Step 1. Access the MAT: https://explorer.ooni.org/chart/mat
Step 2. Adjust the filters depending on the question you would like to answer. For example, to check the recent testing of Twitter in Russia, select Russia in the Country filter, and type twitter.com in the Domain field (as illustrated below).
Step 3. Click Show Chart to plot a chart based on the selected filters.
You will now see a chart with aggregate OONI data, generated based on the selected filters. In the above example, we can see all OONI measurements from the testing of twitter.com on all tested networks in Russia between 12th February 2023 to 11th March 2023. To download the generated data (in JSON or CSV format), click JSON Data or CSV Data at the bottom right corner of the chart.
From the above MAT chart, we can see that most measurements from the testing of twitter.com in Russia presented anomalies (annotated in orange), providing a strong signal of blocking. We can also see that in some cases, the blocking of twitter.com was automatically confirmed (annotated in red).
Step 4. Click on a bar in the MAT chart.
This will show a tooltip with summary information based on the selected measurements. In this example, we can see that on 3rd March 2023: 655 measurements presented anomalies, 56 measurements confirmed the blocking of twitter.com, 86 measurements failed, while only 69 measurements showed that twitter.com was accessible.
Step 5. To access these measurements, click View measurements in the tooltip. This will open the relevant OONI Explorer Search page, pre-filtered to only show the measurements from the selected MAT chart bar.
You can access each of these measurements by clicking on each measurement row, as described in the previous OONI Explorer Search Tool sections of this guide.
Filtering measurements based on multiple domains
If you’re interested in checking the blocking of specific domains in a country, you can do that through the following steps:
Step 1. Select a country (e.g Russia) in the Country filter.
Step 2. Select Domain under Rows.
Step 3. Click Show Chart.
Step 4. Type the domain (www.facebook.com) that you would like to find measurements for in the Search field of the measurement table.
Note: When you start typing the domain, other relevant domains may appear. For example, when typing facebook, you will also see other Facebook domains (such as edge-mqtt.facebook.com and staticxx.facebook.com), in addition to www.facebook.com.
Step 5. Click on the box for the selected domain.
Step 6. Repeat the last 2 steps to add as many domains as you like.
Note: If you don’t find a domain, that’s either because it hasn’t been tested during the selected date range, or because it’s not being tested by OONI Probe users in that country (in which case, you can contribute that website for testing).
Step 7. Click Apply to generate charts based on the selected domains.
Other MAT use cases
Beyond the previous examples, you can use the MAT to plot charts for a variety of different questions. These include:
Checking in which countries a specific website is blocked. For example, if you would like to check the testing of BBC globally:
- Type
www.bbc.com under Domain - Select
Countries under Rows - Click
Show Chart- Click
Confirmed Count in the measurement table - Select countries with the largest volume of
Confirmed Count measurements - Click
Apply
Checking which categories of websites are blocked in a country. For example, if you would like to check which categories of websites are blocked in Italy:
- Select
Italy under Country - Select
Website Categories under Rows - Click
Show Chart- Click
Confirmed Count in the measurement table - Select categories with the largest volume of
Confirmed Count measurements - Click
Apply
Checking which countries block specific categories of websites. For example, if you would like to check which countries block news media websites:
- Select
All Countries under Country - Select
News Media under Website Categories - Select
Countries under Rows - Click
Show Chart- Click
Confirmed Count in the measurement table - Select countries with the largest volume of
Confirmed Count measurements - Click
Apply
Checking which countries block instant messaging apps. For example, if you would like to check which countries are blocking WhatsApp (based on recent measurements):
- Select
All Countries under Country - Select
WhatsApp Test under Test Name - Select
Countries under Rows - Click
Show Chart- Click
Anomaly Count in the measurement table - Select countries with the largest
Anomaly Count measurements - Click
Apply
Comparing censorship across networks. For example, if you would like to check which ISPs in Russia block access to Facebook Messenger:
- Select
Russia under Country - Select
Facebook Messenger Test under Test Name - Select
ASN under Rows - Click
Show Chart
We hope you will find the MAT useful to your research and/or advocacy!
We thank all OONI Probe users worldwide who have contributed – and continue to contribute – measurements, shedding light on internet censorship.
If you have any questions or need data analysis support for your research or investigation, please feel encouraged to contact the OONI team. Thank you for reading this guide!