Decoding content taxonomy in video for industry-specific contextual advertising

The exponential growth of free video platforms is redefining content consumption and its advertising dynamics. Beyond this, the progressive irrelevance of cookies really underlines the necessity for finding a genuinely new and proper way of recording and analyzing user behavior, with an implication for putting user privacy first by design, compliant with data protection and regulatory guidelines. As a result, contextual advertising has become a suitable strategy for the delivery of relevant and personalized advertisements to users. In light of this evolving process, this article proposes a novel framework to facilitate the discovery of contextual information on video content within the industrial context. The proposed framework seamlessly integrates the visual and audio features that are extracted from video content to obtain a comprehensive comprehension of videos. This method optimizes the presentation of ads within a context using a combination of multimodal analysis, industry taxonomy, and contextual advertising. The effectiveness of the framework is validated through experimental results using the YouTube-8M data set, demonstrating its potential to revolutionize contextual advertising by capturing the essence of video content and aligning it with the industry content taxonomy.