Media names matching and normalization

A method comprising:

• receiving a first media data associated with a first exposure via a first media platform, wherein the first exposure is associated with a first media name and associated first metadata;
• receiving a second media data associated with a second exposure via a second media platform, different than the first media platform, wherein the second exposure is associated with a second media name and associated second metadata, the second media name different than the first media name;
• using a machine learning model to determine, based on:
• (i) a relationship between the first media name and the second media name and (ii) a relationship between the associated first metadata and the associated second metadata, that the first media name and the second media name each map to a common, normalized media name corresponding to underlying media content, wherein using the machine learning model to determine that the first media name and the second media name each map to the common, normalized media name comprises:
• generating the common, normalized media name based on a combination of respective weighted scores associated with (i) the relationship between the first media name and the second media name and (ii) the relationship between the associated first metadata and the associated second metadata;
• determining an audience measurement rating for the underlying media content based on the first exposure via the first media platform and the second exposure via the second media platform, each of the first exposure via the first media platform and the second exposure via the second media platform mapping to the common, normalized media name;
• and
• outputting the common, normalized media name in association with reporting of the audience measurement rating.

An at least one non-transitory computer-readable storage medium, having stored thereon program instructions that, upon execution by a processor, cause performance of a set of operations comprising:

• receiving a first media data associated with a first exposure via a first media platform, wherein the first exposure is associated with a first media name and associated first metadata;
• receiving a second media data associated with a second exposure via a second media platform, different than the first media platform, wherein the second exposure is associated with a second media name and associated second metadata, the second media name different than the first media name;
• using a machine learning model to determine, based on:
• (i) a relationship between the first media name and the second media name and (ii) a relationship between the associated first metadata and the associated second metadata, that the first media name and the second media name each map to a common, normalized media name corresponding to underlying media content, wherein using the machine learning model to determine that the first media name and the second media name each map to the common, normalized media name comprises:
• generating the common, normalized media name based on a combination of respective weighted scores associated with (i) the relationship between the first media name and the second media name and (ii) the relationship between the associated first metadata and the associated second metadata;
• determining an audience measurement rating for the underlying media content based on the first exposure via the first media platform and the second exposure via the second media platform, each of the first exposure via the first media platform and the second exposure via the second media platform mapping to the common, normalized media name;
• and
• outputting the common, normalized media name in association with reporting of the audience measurement rating.