Methods, apparatus, and systems to collect audience measurement data

A computing system of an audience measurement entity, the computing system comprising:

• a memory;
• and
• one or more processors configured to:
• obtain audience member behavior data developed based on first program identification data and audience identification data, the first program identification data and the audience identification data collected from a first household,
• obtain, from a media service provider, second program identification data for a television of a second household that is different from the first household, wherein the second household is a non-metered household for which the second program identification data is obtained without using a meter to collect audience identification data at the second household,
• determine, for an audience member of the second household, using a model that is trained with the audience member behavior data, an audience-member viewing probability for television content identified in the second program identification data, the audience-member viewing probability representing a likelihood that the audience member viewed the television content, and
• generate a viewership report for the television content using the audience-member viewing probability and demographic information for the audience member,
• wherein the computing system of the audience measurement entity is located remotely from the first household and the second household.

A method comprising:

• obtaining audience member behavior data developed based on first program identification data and audience identification data, the first program identification data and the audience identification data collected from a first household;
• obtaining, from a media service provider, second program identification data for a television of a second household that is different from the first household, wherein the second household is a non-metered household for which the second program identification data is obtained without using a meter to collect audience identification data at the second household;
• determining, by a computing system of an audience measurement entity, for an audience member of the second household, using a model that is trained with the audience member behavior data, an audience-member viewing probability for television content identified in the second program identification data, the audience-member viewing probability representing a likelihood that the audience member viewed the television content;
• and
• generating a viewership report for the television content using the audience-member viewing probability and demographic information for the audience member,
• wherein the computing system of the audience measurement entity is located remotely from the first household and the second household.