Abstract
Native advertisements are advertisements that closely match the look-and-feel of the publication in which the ads are placed and are an important category of online advertisements. It is important for a publisher to determine whether a given native ad is of high quality and adheres to a stylebook for the publication. Manual evaluation of native ads can be costly, time consuming, and subjective. This disclosure describes use of machine-learning techniques to automatically score a native ad style for aesthetics and conformance to stylebooks. Publishers benefit with the automated quality assessment of a given native ad style and can improve aesthetics and monetization of native ads. Ad-buyers can adjust their bids based on the quality assessment.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Lin, Bo and Fu, Bin, "Automatic quality assessment of native advertisements", Technical Disclosure Commons, (March 30, 2018)
https://www.tdcommons.org/dpubs_series/1119