This is an automated email from the ASF dual-hosted git repository. spmallette pushed a commit to branch TINKERPOP-3158 in repository https://gitbox.apache.org/repos/asf/tinkerpop.git
commit 6362002e35c9bbbd8cec7c6415a2fb5f230cfeea Author: Stephen Mallette <[email protected]> AuthorDate: Tue Jun 2 08:36:37 2026 -0400 Moved Tinker vector search index to 4.0.0 --- docs/src/upgrade/release-4.x.x.asciidoc | 45 +++++++++++++++++++++++++++++++++ 1 file changed, 45 insertions(+) diff --git a/docs/src/upgrade/release-4.x.x.asciidoc b/docs/src/upgrade/release-4.x.x.asciidoc index 2b4606444c..152fdd8cbf 100644 --- a/docs/src/upgrade/release-4.x.x.asciidoc +++ b/docs/src/upgrade/release-4.x.x.asciidoc @@ -41,6 +41,51 @@ anonymized form. The original gremlator.com was a prototype built by TinkerPop c previous implementation required Java and a running Gremlin Server, whereas the new version runs entirely in the browser with no server infrastructure needed. +==== TinkerGraph Vector Index + +Vector search enables finding similar items based on their semantic meaning rather than exact keyword matches, which is +essential for modern applications like recommendation systems, image search, and natural language processing. +TinkerGraph now supports vector indexing and similarity search through a dedicated vector index implementation and the +`call()` step, allowing efficient nearest neighbor searches on both vertices and edges. + +The vector index in TinkerGraph is implemented using HNSW (Hierarchical Navigable Small World), a state-of-the-art +algorithm for approximate nearest neighbor search that provides excellent performance for high-dimensional vector data. +To use this feature, you need to create a vector index, add elements with vector embeddings, and then perform searches +using the `call()` step. + +Here's a basic example using Groovy: + +[source,groovy] +---- +// Create a graph and register the vector search service +graph = TinkerGraph.open() +g = traversal().with(graph) +graph.getServiceRegistry().registerService(new TinkerVectorSearchByElementFactory(graph)) + +// Create a vector index with dimension 3 +indexConfig = [dimension: 3] +graph.createIndex(TinkerIndexType.VECTOR, "embedding", Vertex.class, indexConfig) + +// Add vertices with vector embeddings +g.addV("person").property("name", "Alice").property("embedding", new float[]{1.0f, 0.0f, 0.0f}).iterate() +g.addV("person").property("name", "Bob").property("embedding", new float[]{0.0f, 1.0f, 0.0f}).iterate() +g.addV("person").property("name", "Charlie").property("embedding", new float[]{0.0f, 0.0f, 1.0f}).iterate() + +// Perform a vector search using the call() step +params = [key: "embedding", topK: 2] +g.V().has("name", "Alice").call("tinker.search.vector.topKByElement", params) +---- + +The `call()` step returns a list of maps, each containing the matching element and its distance from the query vector: + +[source,text] +---- +==>[distance:0.9500624, element:v[2]] +==>[distance:1.0, element:v[3]] +---- + +See: link:https://issues.apache.org/jira/browse/TINKERPOP-3158[TINKERPOP-3158] + ==== Removed `uuid` Dependency in gremlin-javascript The `uuid` npm package has been removed from `gremlin-javascript`. UUID generation now uses the built-in
