How is Machine Learning used in the LinkedIn Recruiter Recommendation System

mediumThis post was originally published by Jitendra Singh Balla at Medium [AI]

Image for post
Photo by inlytics on Unsplash

Let’s find out!

Primary reason LinkedIn users are active on the platform for job recruitment efforts. With more than 20 million companies listed on the site and 14 million open jobs, it’s no surprise to find out that 90% of recruiters regularly use LinkedIn.

In fact, a study found that 122 million people received an interview through LinkedIn, with 35.5 million having been hired by a person they connected on the site.

Heavy Usage of ML and DS

Recruiter Recommendation

This product by LinkedIn needs to handle arbitrarily complex queries and filters and deliver results that are relevant to specific criteria.

The Architecture

  • The inverted field: a mapping from search terms to the list of entities (members) that contain them​.
  • The forward field: a mapping from entities (members) to metadata about them.​

These search index fields contribute to the evaluation of machine learning feature values in search ranking. The freshness of data in the search index fields is also of high importance for machine learning features.

Image for post

Source : https://engineering.linkedin.com/blog/2019/04/ai-behind-linkedin-recruiter-search-and-recommendation-systems

The Ranking Model

  • L1: Scoops into the talent pool and scores/ranks candidates. In this layer, candidate retrieval and ranking are done in a distributed fashion.​
  • L2: Refines the short-listed talent to apply more dynamic features using external caches.​

Image for post

Source: https://engineering.linkedin.com/blog/2019/04/ai-behind-linkedin-recruiter-search-and-recommendation-systems

The Details

  • The Galene broker system fans out the search query request to multiple search index partitions.
  • Each partition retrieves the matched documents and applies the machine learning model to retrieved candidates.
  • Each partition ranks a subset of candidates, then the broker gathers the ranked candidates and returns them to the federator.
  • The federator further ranks the retrieved candidates using additional ranking features that are dynamic or referred to from the cache — this is the L2 ranking layer.

Finding a Good Fit

Image for post

Machine Learning Methodologies used

  • Context-aware ranking with pairwise learning-to-rank.
  • Deep and representation learning.
  • Large-scale information Network Embedding.
  • Entry-level personalization with Generalized Linear Mix (GLMix).

If you want to learn Exploratory Data Analysis (EDA) in easiest way using Python than click here.

If you want to know about NumPy library of Python than just click here.

Spread the word

This post was originally published by Jitendra Singh Balla at Medium [AI]

Related posts