This Policy was last modified on August 4, 2017
- 1. Overview
- 2. Applicability and Notification
- 3. Consent and Revocation
- 4. Collection of Data
- 5. Data Storage and Control
- 6. Use of Data
- 7. Contact
Measurement Lab is an international consortium that provides one of the largest collections of open data on Internet performance, all generated by users that voluntarily conduct experiments. Through these experiments, M-Lab collects data that provide crucial insights into the current state of the Internet for the public, researchers, technical communities, businesses, and policymakers.
M-Lab recognizes that the security and privacy of collected data are important concerns for users. The data minimization practices of the platform are governed under a separate Acceptable Use Policy (“AUP”) that requires that M-Lab and its partners only collect data necessary to understand Internet performance. Since all data collected through experiments is released publicly, M-Lab’s AUP requires the submission of only specific technical Internet measurements and restricts the collection of sensitive personal information, such as physical location or web browser history.
If you have any questions about this Policy, please contact us at: firstname.lastname@example.org
2. Applicability and Notification
This Policy explains M-Lab’s practices concerning the collection, storage, and publication of data that occurs in connection with the operation of its services.
The M-Lab platform may be utilized or included within software or hardware clients that are developed and maintained by third parties. The utilization of the M-Lab platform by partners is governed by our AUP, which sets out expectations on data collection and privacy. Partners may apply their own privacy policies or terms of service to the use of their client or experiment, and may engage in additional data collection beyond the scope of data collected through the M-Lab services, provided that the conditions set forth in the AUP are met. Information pertaining to partners or third parties involved in the provision of a client should be provided to the user per the M-Lab AUP. For questions or information on the policies and data collection of specific experiments or clients, please refer to the entity responsible for that experiment or client (as indicated in this List of Experiments).
M-Lab may modify this Policy from time to time. We will publish updates to this Policy on our blog and send notifications to the M-Lab discussion list.
3. Consent and Revocation
Any data collected in the use of M-Lab services will remain stored in M-Lab’s dataset after collection is stopped. Since M-Lab does not collect personal information, it does not have the ability to delete collected data in response to requests. Additionally, M-Lab immediately publishes data upon collection, and therefore cannot rescind data placed into the public domain or remove data from third party copies of the dataset.
4. Collection of Data
M-Lab publishes datasets on the performance of the Internet as it is experienced by users around the world. In order to do so, users utilize clients to conduct experiments hosted on the M-Lab platform from their device, providing data that characterizes the connection between the user and an M-Lab server. While the specific data collected differs across experiments and clients, M-Lab broadly stores three types of data: measurement data, connection data, and metadata. For more about the methods followed and data collected by different experiments, see the M-Lab Tests page.
Partners who develop experiments and clients that use M-Lab’s platform are all bound by M-Lab’s AUP by their use of the platform. Partners are solely responsible for ensuring their compliance with the AUP.
M-Lab’s AUP does not permit collection of information about content of communications, such as emails or Internet searches. The experiments hosted on the M-Lab platform generate artificial traffic rather than use real user traffic based on AUP requirements, in order to avoid collection of sensitive information. To ensure the integrity of the platform for scientific research, M-Lab does not take additional measures to sanitize or alter the raw data.
a. Measurement Data
The technical metrics and measurements collected for each experiment differ. Most experiments measure performance variables describing the state of connection during the test, commonly known as ‘key performance indicators’ or ‘quality of service indicators.’
These indicators include, for example:
- Average download and upload speed;
- Latency and topology of the path between the client and server; and,
- Whether the application or protocol is manipulated or discriminated against while in transit.
b. Connection Data
M-Lab collects and publishes the:
- public-facing Internet Protocol (IP) address of the client that conducted the measurement; and
- the date and time of the test.
The IP address and time of the test are necessary to understand and describe experiment results.
Subject to the AUP, clients or experiments may collect additional descriptive information about the user’s environment in order to support analysis of measurement data. The decision whether to provide this data within test results is left to the discretion of the partners. M-Lab’s AUP does not allow the storage of information on the platform that would associate individuals or clients to their history of tests or devices, such as the hardware identifier, service plan, or network signal metrics.
Metadata collected by a client may include:
- the name and version of the web browser software;
- the software application; and,
- the operating system used to conduct the test.
5. Data Storage and Control
M-Lab’s platform is internationally distributed and dependent on cloud services, which has implications for the collection and storage of data.
The initial point of collection is the server on the M-Lab platform with which a client conducts a test. Clients are not restricted in the servers they may use. For a single user, the initial point of collection can differ between measurements and may or may not be in the user’s country of residence.
After collection, M-Lab aggregates the data from the servers on the platform. This process enables the long-term storage and analysis of the data. The primary processing pipeline used by M-Lab for this task makes the data available on Google’s BigQuery and Cloud Storage services (subject to the Google Cloud Terms of Service). Data hosted through these services may be distributed across multiple countries.
The data is published under a Creative Commons Public Domain Dedication (CC0 1.0) license upon collection. Herewith M-Lab waives all copyright and related rights in and to the data to fullest extent allowed by law. Since the data is openly licensed and published upon collection, M-Lab cannot restrict the jurisdictions in which data are distributed or accessed after publication.
6. Use of Data
M-Lab releases all of its data into the public domain, free of any use restrictions, to ensure that everyone is free to use the open dataset. We encourage the use of M-Lab data in order to support research and public learning.
M-Lab uses the data for the following purposes:
- To provide the services and to support the process of conducting experiments, such as directing a client to an appropriate server on the platform;
- To aggregate data in its open dataset; and,
- To support transparency and usability initiatives, such as the production of network performance visualizations.
M-Lab, its partners, and other third parties may also use the data for research and other projects, illustrative examples of which are detailed on our publications page. The data is used for diverse purposes, and the choice of license is designed to enable research without restrictions. Third parties may use the data for any purpose, without needing to seek prior permission from M-Lab. M-Lab cannot assert control over third parties’ use of the published data.
|Platform||M-Lab operates a large fleet of servers based on open source software that are distributed around the world in order to support research on the Internet.|
|Experiments||M-Lab’s platform hosts measurement projects, generally developed by third parties, such as academic institutions. These experiments are designed to utilize technical methods to collect data on specific aspects of Internet performance. While M-Lab imposes strict policies on what the experiments can or cannot do with the platform, the experiments are maintained by their creators, who are responsible for Policy compliance. Information about the experiments can be found in detail on the page for each Experiment.|
|Clients||The software application or hardware equipment that tests the connection between the user and an M-Lab server. The actual form of the client can differ, such as being a mobile application or a network probe that conducts experiments. Since M-Lab is an open platform, the maintainer of the experiment is not necessarily the same entity as the developer of the client, and a number of experiments have third-party clients.|
|Collection||The process wherein a client conducts an experiment on the platform, and the resulting data is retained on a M-Lab server.|
|Storage||The process of aggregating the results of collected data for experiments on all servers within the M-Lab platform, and retaining those results for indefinite long-term archival and data processing purposes.|
|Publication||The release of measurement results within the public dataset that is made available by M-Lab.|
|Partners||Individuals or entities that develop or maintain experiments or clients that use the M-Lab platform. M-Lab does not necessary maintain a legal or operational relationship with such entities outside of the provision of platform resources under the AUP.|