LinkedIn's search functionality is one of its oldest capabilities, allowing members to search for people they know, or to discover new connections. With 900 million members on the platform, LinkedIn faces the challenging task for every search you make: who are the most relevant people to show you?
Unlike other recommendation systems, machine learning systems contain one key component not found elsewhere: the query. User intent is textually explicit through the query, and search results must conform accordingly.
In this presentation, we'll explore how LinkedIn's People Search system uses ML to surface the right person that you're looking for, including but not limited to:
1. Retrieval: determining the profiles relevant to your search intent
2. Ranking: selecting the most relevant profiles to show you
Expertise in ML is not a pre-requisite to enjoy this presentation, and while some background is helpful, all are encouraged to attend.
Machine Learning @ LinkedIn
Mat is passionate about helping others find what they need.
As an ML engineer at LinkedIn, he leads the technical development of the ML behind People Search, LinkedIn's search engine that helps members find other people that are interesting to them. Before that, he built the NLP system at brain.ai, an early-stage startup that helps users shop by simply saying what they need. Before that, he worked as a Data Scientist at Quora, analyzing experiments that helped users find answers to their questions.
When he is not finding local minima in high-dimensional spaces, Mat enjoys finding local minima in his ski boots, or local maxima in his hiking boots.