Nathan is a Senior Data Scientist at Neo4j. He uses data and algorithms to solve problems and demystifies machine learning to help people use the tools of data science to make the world a better place.
Example graph showing User nodes related to Email nodesGraph databases like Neo4j are great for solving entity resolution problems. For example, a graph shows where two user profiles share a common identifier, like an email address. However, it can be challenging for a graph to identify cases... read more
Semantic search allows search systems to retrieve documents that match the meaning of a query even if the exact keywords in the query are not present in the document. This flexible retrieval capability is a key part of many Retrieval Augmented Generation (RAG) applications in Generative AI.RAG... read more
In a previous article, I discussed the benefits of using k-medoids to cluster graph data. In the k-medoids approach, you determine how many clusters you would like to partition the graph into. This number is called k. The algorithm identifies a set of k nodes in the graph called medoids. The other... read more
K-medoids is an approach for discovering clusters in data. It is similar to the well-known k-means algorithm.Both approaches require the analyst to select the number of output clusters before running the algorithm. This number is called k. Both algorithms assign each dataset member to one of... read more
Cluster related products and separate conflicting entities with the newest algorithm in the Graph Data Science Library: Approximate Maximum K-cutPhoto by Matt Artz on UnsplashThe 1.7 release of Neo4j’s Graph Data Science Library contains some amazing features, like machine learning... read more
