LinkedIn is set to update its post-search architecture, which will produce quicker and more pertinent results.
To improve the relevance of search results, LinkedIn set out to create a system that considers the following elements: the post’s relevance to the query, its quality, personalization, user intent, interaction, and its freshness/recency.
LinkedIn claims on their blog post on August 25, 2022 that the new system must also deliver data from a variety of sources and to accomplish these objectives for relevance and diversity, however, LinkedIn said they are using several machine learning techniques to match searcher expectations.
Furthermore, LinkedIn used data from crowdsourcing human reviews of search results to make sure its new system passes a certain bar for quality. They added that the data collected through crowdsourced human annotations also serve as useful training data to raise the ranking of results.