In this lecture, Professor Jure Leskovec delves into the process of capturing data, modeling patterns, and predicting behavior using vast amounts of textual data gathered from blog posts and news articles. With access to over 20 million pieces of content daily, he outlines methods for mining this data to identify temporal patterns in news coverage on a daily basis. Specifically, he focuses on analyzing the evolution and competition of storylines as they gain and lose attention over time.

Moreover, Professor Leskovec discusses models designed to quantify the impact of individual media outlets on the popularity of news stories. He explores algorithms capable of uncovering hidden networks that facilitate the flow of information, shedding light on the complex dynamics underlying the dissemination of news. By understanding these patterns and networks, researchers can gain insights into how information spreads through various channels, potentially influencing public perception and behavior.

Overall, this lecture provides valuable insights into the intricate processes involved in analyzing large-scale textual data, offering strategies for capturing, modeling, and predicting patterns of information dissemination in the media landscape