Capturing Signals to Personalize Context
Fusion captures and aggregates signals like queries, clicks, views, and purchases to create a customized, optimized experience for each user.
Relevancy Tuning
Relevancy tuning, based on a user’s context, provides a flexible experimentation framework that adjusts results according to changes in user behavior.
Boosting with Signals
Query pipelines utilize signals (such as clicks or page views) to selectively boost items in the set of search results. As signals accumulate, the aggregate becomes more predictive.
Learn to Rank (LTR)
LTR extracts tags such as product names, titles, document categories (e.g., marketing, support, architecture diagrams) to determine relevance scores in real-time.