BuzzFeed’s mission is to share the best content of the internet with the world through the hundreds of original articles, videos, and graphics it produces daily. Its team of human curators handpick the content that is most relevant and interesting to audiences across its vast network of social media distribution channels. Of course, as content production scales and social networks grow, curation becomes increasingly difficult.
To this end, the company has built a curation workflow powered by natural language processing and deep learning techniques. It learns to find the most relevant content for each channel, identify and resurface content with evergreen appeal, and determine the optimal times in the day to post each content. Join Lucy Wang and Swara Kantaria to explore real-world examples of how neural networks helped boost performance at BuzzFeed. They discuss the implications of each use case, explaining how the company developed its methods and how the service cohabitates with existing human curation in its internal workflow tools.
Lucy X. Wang is a senior data scientist at BuzzFeed working on machine learning tools for optimizing audience reach and engagement. She performed research on social networks and information diffusion at Columbia University.
Swara Kantaria is a senior product manager at BuzzFeed, where she’s worked on a range of products from launching Buzzfeed.com in international markets to building new internal tools. Swara is the product lead for distribution tools, a team responsible for building products that send all of the company’s content to social media platforms (like Facebook, YouTube, Twitter, and Instagram), OTT destinations (like Roku and Pluto), and syndication partners. Previously, Swara was a technology consultant at Deloitte.
Comments on this page are now closed.
©2019, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • email@example.com