Evaluation of streaming video content material and era related contextual promoting



Tetiana Kovaliuk (1), Nataliya Kobets (2), Grigorii Shekhet (3) and Tamara Tielysheva (3)

1 – Taras Shevchenko Nationwide College of Kyiv, 64/13, Volodymyrska str, Kyiv, 01601 Ukraine
2 – Borys Grinchenko Kyiv College, 18/2 Bulvarno-Kudriavska Str, Kyiv, 04053, Ukraine
3 – Nationwide Technical College of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, 37, Prospekt Peremohy, Kyiv 03056, Ukraine

The speech discusses the duty of looking and deciding on ads that match the content material of the video. The issue is pressing as a result of great amount of video and promoting content material on the Web. The authors set the duty to develop info know-how to resolve this drawback. The strategies of recognizing the contours of objects utilizing the Prewitt operator and looking for comparable objects in streaming video primarily based on perceptual hash capabilities are thought-about. An algorithm for recognizing speech from an audio monitor primarily based on Mel-Frequency Cepstral Coefficients and an algorithm for looking for key phrases within the textual content from a sound monitor primarily based on TF-IDF are developed. The thought-about algorithms are used to find out contextual promoting for video content material. The introduced management instance, the outcomes of which point out the effectivity of the algorithm.

Leave a Reply

Your email address will not be published. Required fields are marked *