In a world inundated with data, AI and machine learning are indispensable for making sense of the data we have at scale. At LinkedIn, we’ve built the Economic Graph, a digital representation of the global economy made up of over 660 million members, 36 thousand skills, 30 million companies, 20 million open jobs, and 90 thousand schools. To use this data to connect people to economic opportunity, AI is woven into our platform to improve every aspect of the member experience. Whether you want to look for your next play, expand your network, or simply stay informed on your industry, AI makes these experiences more personalized, powerful, and relevant. In this blog, I’ll share three examples of how AI is used at LinkedIn to help the world’s professionals be more productive and successful.
Creating equal opportunity for job-seekers
For recruiters and hiring managers, our Talent Search systems play an important part in connecting candidates with roles by sifting through members to identify “talent pools” for a specific opportunity. However, examples of the unintended effects of bias in algorithmic systems and research on how members represent themselves online point towards the need to proactively prevent real-world bias from creeping in our recommendation systems. Our ranking algorithms not only take into consideration a particular candidate’s interest in an employer and their preferred location to work, but also measure representativeness to return search results that will help our recruiter customers source diverse talent. This means that the top search results for a given role will be representative of the broader qualified candidate set, helping us create opportunity for everyone—not just a select few.
Surfacing a holistic balance of content
Every day, the LinkedIn feed ingests over a million posts, updates, videos, and articles. Viewing the feed as an ecosystem, where the goal is to optimize the experiences of both content creators and viewers, has boosted engagement on both sides. This is done by ensuring that our feed relevance models are tasked with not only serving content relevant to a member’s interests, but also considering nuances, such as the experiences of the creator and viewer’s network. For example, feeds focused solely on engagement on the viewer side will develop a preference to display a majority of content from influencers and celebrities.
A more holistically optimized feed that takes into consideration multiple objectives—value to the member, member’s network, and creator—for recommending content. It’s not just about virality, but about considering how a single act of engagement can be meaningful for our creators who are not necessarily trending, but are creating high-quality content. In implementing this approach into our feed model, we saw wins for both creators and viewers—members liked seeing more content from people they know, while creators were further incentivized to share more on the platform, bringing a better experience for all
A core component of building communities is connecting individuals through common interests or shared experiences. To build this experience digitally, we use AI to help members discover, engage, and contribute. It starts with determining a member’s interests with our Follow Recommendations product, which presents members with schools, companies, or groups that the member might find relevant and engaging in following.
Once a member subscribes to an entity, related content will populate their feed for them to engage with. To help members contribute back to their communities, we also recommend relevant hashtags and typeaheads that will help them effectively target their posts to the right audiences. For example, if you begin typing a post on deep learning, we will suggest you add “#AI” or “#artificialintelligence”. By weaving AI into these three different components, we’re committed to fostering productive conversations and active communities on our platform.
'To learn more about LinkedIn's experiences in building a data-first culture, come and hear Deepak delivering a keynote address at World Summit AI Americas on 25th-26th March 2020, Montreal'.
GLOBAL AI EVENTS CALENDAR
Here is your Global AI Events Calendar where you can meet the Inspired Minds community of business leaders, heads of government, policy makers, startups, investors, academics and media.
INTELLIGENT HEALTH UK
5th February 2020
ExCeL London, UK
WORLD SUMMIT AI AMERICAS
March 25-26th 2020
Palais des congrès de Montréal
09-10 September 2020
WORLD SUMMIT AI
07-08 October 2020
Amsterdam, The Netherlands