Jun XieSemantic Search ETLETL for semantic search (extract, transform and load) aims to retrieve data from different systems, transform the raw data (e.g., text…Oct 6, 2022Oct 6, 2022
Jun XieTop Pre-trained Models for Sentence EmbeddingThis article walks through top pre-trained models to get sentence embedding, which is a lower-dimensional numerical representation of the…Oct 3, 2022Oct 3, 2022
Jun XieWeb-Scale Recommendation SystemsThis article walks through a tutorial on web-scale recommendation systems, which are normally built in the largest Internet companies…Oct 3, 2022Oct 3, 2022
Jun XieSemantic search on billion-level documentsSharding is needed to enable semantic search on billion-level documents. If the semantic search traffic is high, we also need to do the…Sep 13, 2022Sep 13, 2022
Jun XieTop 5 Pre-trained Model for Image EmbeddingPre-trained models help boost the popularity of semantic search. We can easily get embedding (aka. vector) of different media (i.e. text…Sep 1, 2022Sep 1, 2022
Jun XieSemantic Search KNNSemantic search uses K-nearest neighbor search (KNN) to return the most relevant results given a query (i.e. vector + metadata filtering)…Aug 30, 2022Aug 30, 2022
Jun XieSemantic Search ANNApproximate nearest neighbor search (ANN) makes semantic search possible and blazing fast on billions of documents. ANN is also called…Aug 30, 2022Aug 30, 2022
Jun XieSemantic Search ApplicationsSemantic search powers more and more industrial applications, including recommendation, and document/image/audio/video retrieval.Aug 27, 2022Aug 27, 2022
Jun XieSemantic Search Similarity MetricsThere are three common similarity metrics used in the semantic search to measure the similarity/distance between vectors, including squared…Aug 26, 2022Aug 26, 2022
Jun XieRelated Product RecommendationThis article walks through a general architecture to support the related/similar product recommendation based on semantic search.Aug 25, 2022Aug 25, 2022