Predicate Proposal - Phone Number

Value statement: Phone numbers contain relevant contact information for an entity.

Definition: “The phone number associated with this entity that is 1. public (it is openly known the entity is associated with this phone number) and 2. official (the entity has formally associated themselves with this phone number by posting it on a website, social media link, etc. that they are in control of)”.

Tooltip definition of the predicate: “The public and official phone number associated with this entity”

Type of value: String

For enums - all possible values: N/A

# of accepted values: multiple - i.e. some users will maintain multiple phone numbers, particularly companies/organizations.

Inverse Properties and Name: N/A

Examples of proper use:
- ‘Apple’ ‘Phone number’ is ‘+1-800-275-2273‘ citation: https://support.apple.com/en-us/HT201232
- ‘White House’ ‘Phone Number’ is ‘+1-202-456-1414’ citation: https://www.whitehouse.gov/get-involved/write-or-call/

Examples of improper use:
- ‘Arianna Grande’ ‘Phone Number’ is ‘x-xxx-xxx-xxxx’, where ‘x-xxx-xxx-xxxx’ is not a publicly disclosed number

Usage in other schemas:
- Wikidata https://www.wikidata.org/wiki/Property:P1329
- Schema.org https://schema.org/telephone

Constraints: Should only be applied to
- companies/organizations
- people

Restrictions (ie PII concerns): there may be PII concerns associated with a phone number. It seems best to handle these one-off - i.e. remove the value only if someone requests that it be removed, which is complicated but possible once this data is stored in IPFS.

Suggested applicable templates on golden.com: company, organization, port, laboratory, educational institution

Also known as: telephone number, landline number, mobile number

Estimated cardinality of this predicate:
- ~4 billion phone numbers for entities that would reasonably be in Golden’s graph
- Assume 1 billion people will be in Golden’s graph. Assume a person has, on average, 1 phone number associated with them (many people will not have a phone number, while some might have more than one)
- Assume 1 billion companies/organizations will be in Golden’s graph. Assume a company/organization has, on average, 3 phone numbers associated with it (large companies/organizations may have multiple phone numbers)

Estimated frequency of new values or changes: Phone numbers are changed very infrequently, by either a company/organization or a person. Companies/organizations may introduce a new phone number on average once every 2-3 years.

Very minor question here. We say only companies/organizations/people here, but in "Suggested applicable templates below we say: “company, organization, port, laboratory, educational institution, person”

Do we want to expand the constraints to include port, laboratory, educational institution?

It is necessary to match addresses and phone numbers, otherwise it is difficult to figure out when several addresses and phone numbers

It is also correct to bind e-mail to an address. The final form should be the following Address, email, fax phone. Now these are separate displayed and non-comparable blocks - that is, in the presented Golden information it is very difficult to correctly match the address-e-mail.-contact telephone-fax

Noted on this and the previous comment. I think this is better addressed with a separate, specific proposal around how we could potentially bundle two predicates together.

This predicate has been accepted into the protocol schema - proposed adjustments to the definition should take place in new forum posts.

1 Like