You are here
Background and Goals
The WHOIS ARS project was created both in response to recommendations compiled and delivered by the 2012 WHOIS Review Team, under the Affirmation of Commitments (AoC), as well as to address GAC concerns on WHOIS accuracy. ICANN committed to proactively identify potentially inaccurate gTLD WHOIS contact data and forward this information to gTLD Registrars for investigation and follow-up.
WHOIS ARS Phases
The ARS is divided into three phases based on the types of validation identified in SAC058:
- Phase 1: Syntax Accuracy
- Phase 2: Syntax + Operability Accuracy
- Phase 3: Syntax + Operability + Identity (TBD; requires further consultation with the community as to if and how this phase would be implemented)
Accuracy Testing Methods
Syntax and operability accuracy testing were designed to assess the contact information of a WHOIS record by comparing it to the applicable contractual requirements of the RAA.
- Syntax testing assessed the format of a record (e.g., does the email address contain an “@” symbol?)
- Operability testing assessed the functionality of the information in a record (e.g., did the email not get bounced back?).
- The resulting data were analyzed to produce statistics of syntax and operability accuracy for WHOIS contact information across subgroups such as New gTLDs or Prior gTLDs, Region, and RAA type (i.e., 2009 RAA or 2013 RAA, https://www.icann.org/resources/pages/registrars/registrars-en)
A two-stage sampling method is used on the WHOIS ARS project to provide a large enough sample to reliably estimate subgroups of interest, such as ICANN region, New gTLD or Prior gTLD, and RAA type. Two samples are prepared at the beginning of each report cycle:
- An initial sample of 100,000-200,000 WHOIS records
- A sub-sample of the initial sample of 10,000-12,000 WHOIS records, which is used for accuracy testing
To provide comments on WHOIS ARS or suggestions to improve WHOIS ARS reports and validation criteria, please email: firstname.lastname@example.org.
More Information on Project Goals
On 8 November 2012, the ICANN Board approved a series of improvements to the manner in which ICANN carries out its oversight of the WHOIS Program, in response to recommendations compiled and delivered by the 2012 WHOIS Review Team, under the Affirmation of Commitments (AoC). ICANN committed to proactively identifying potentially inaccurate generic Top Level Domain (gTLD) WHOIS contact data and forwarding potentially inaccurate records to gTLD registrars for investigation and follow-up. To accomplish these tasks and address Governmental Advisory Committee (GAC) concerns on WHOIS accuracy, ICANN initiated the development of the WHOIS Accuracy Reporting System (ARS)—a framework for conducting repeatable assessments of WHOIS accuracy, publicly report the findings, and provide data to the ICANN Contractual Compliance team to follow up on potentially inaccurate records with registrars. With input from the community, ICANN designed the ARS to be organized into three Phases based on the types of validations described in the SAC058 Report (syntax, operability, and identity, see section “WHOIS ARS Phases” below for more information).
More Information on WHOIS ARS Phases
Phase 1 analyzes the syntax accuracy of WHOIS contact information. Phase 2 is ongoing and cyclical and assesses the operability of the contact data in the record by combining the syntax tests from Phase 1 with operability tests. Phase 3 is intended to look at identity validations. ICANN will continue to work with the Community to assess if Phase 3 will be implemented at all and, if so, how.
More Information on Accuracy Testing Methods
Syntax and operability accuracy tests are performed on all nine individual contact information fields in a record (i.e., email address, telephone number, and postal address for the registrant, administrative, and technical contacts) and compiled as an entire record. The accuracy tests are designed in such a way that all records in the analyzed subsample are evaluated against a set of baseline requirements derived from the requirements of the 2009 RAA. While operability requirements differ little between the 2009 and 2013 RAA versions (only that the Registrant email address and telephone number are not required for 2009), the 2013 RAA requires the contact data in a WHOIS record to be more syntactically complete and to be formatted per more specific requirements than that of the 2009 RAA. For example, the 2009 RAA requires postal addresses with a valid country, whereas the 2013 RAA requires the country in the address to be formatted per the 2-letter code from ISO-3166-1 (http://www.iso.org/iso/home/standards/country_codes.htm). More information on the methodology of this study and the accuracy tests performed can be found in the here: https://whois.icann.org/en/whoisars-validation.
More Information on Sample Design
Our analysis includes three mutually exclusive RAA subgroups: 2009 RAA, 2013 RAA GF, and 2013 RAA non-grandfathered (referred to as 2013 RAA NGF). Though an estimated 97 percent of domain names are registered through registrars which have been accredited under the 2013 RAA, a majority of domains are allowed to operate under the WHOIS standards of the 2009 RAA. This could be for one of two reasons: 1) the registrar has not yet signed a 2013 RAA with ICANN and is only subjected to 2009 RAA standards; or, 2) the registrar agreed to 2013 RAA with ICANN but the domain was registered before the effective date of the registrar’s 2013 RAA. We refer to the latter group of domains as 2013 RAA Grandfathered (2013 RAA GF) domains. For this reason, the 2009 RAA criteria is used as the baseline to assess WHOIS accuracy. More detailed information can be found in the individual Sample Design sections of each report.